by Tim Pestell
The ports of the Seine: from abandonment to reoccupation by the Vikings
Transferring a large part of the population of the Seine vici to Rouen around 888–90, led to a massive desertion of these ports. Correlating numerous sources makes us realise the extent of this phenomenon, and the sites concerned may be divided into two categories: those ports which were permanently abandoned or for a long time, and those that were reoccupied by the Scandinavians.
During the exodus of the clergy from the Cotentin to the Rouen region in 889/890, several groups came to settle in the villages on the banks of the Seine vacated by the inhabitants. The best documented case is of the three ports in the Jumièges region, all ancient dependencies of the abbey. According to a late twelfth-century account inserted in a collection, the Miracula Sancti Pauli, the place called Saint-Paul, on the north bank (localité de Saint-Paul, commune Duclair), had been abandoned some time during the 880s. A group of monks reoccupied the houses surrounding the church and placed in the latter the body of St Clair which the clerics had carried with them on leaving the Cotentin. We are less well informed about the conditions prevailing with the setting up of two other groups, one with the body of St Hameltrude at the port of Saint-Vaast, on the south bank opposite Jumièges (loc. de Saint-Vaast, comm. Heurteauville), the other with the body of St Pèlerin in a port on the north bank (loc. de Le Passage, comm. Jumièges). The new occupants only lived there a few years; after their departure, the three sites were abandoned again, leaving only the churches, of which two fell into ruins (loc. de Le Passage and Saint-Paul) (Le Maho 1999, 150–60). The other monastic ports of the Seine shared a similar fate. Those of Fontenelle Abbey, in the Caude-becquet region, were abandoned for good; when in 960 this settlement was restored, the monks’ port was transferred to Caudebec. A tradition of ‘deserted village’ was also attached to the hamlet of Grestain in the Seine bay, probable site of the abbey of Pennante (comm. Saint-Pierre-du-Val, Eure), mentioned for the second time in 833. At the beginning of the eleventh century, only the ruins of the ancient chapel of Notre Dame remained on the site.
It would seem that the cases of prolonged abandonment were in a minority. Indeed, if we examine the map of Scandinavian place-names in the Lower Seine, the evidence is striking, although the density of these place-names is no more significant in this region than elsewhere.
FIGURE 18.1. The ports of the Lower Seine in the ninth and tenth centuries.
But, in all cases, the sites are closely linked to the river economy, as mercantile and harbour villages, and fishing or sail-making hamlets. Moreover, there are numerous signs of occupation on these sites from the pre-Viking period.
We may begin our inventory on the north bank, progressing upstream (Fig. 18.1). At the entrance of the bay, near the point of Chef-de-Caux, is Sanvic, which seems to correspond to the anonymous portus of Saint-Denis Abbey, mentioned in a charter of 810. Next is Harfleur at the mouth of the Lézarde, a port which was perhaps part of the possessions of Montivilliers Abbey in the pre-Viking era, and which would appear to be the ancient Caracotinum of the Antonine Itinerary. Then there are Orcher, Oudalle and Sensedalle, where three small valleys overlook a vast expanse of salt marshes; and Villequier, a possession of Fontenelle Abbey since the seventh century and probably one of the ‘Mesnils’ on the banks of the Seine mentioned in a diploma of Charles the Bald of 854; and Caudebec-en-Caux, once a river port, and crossing point on the Rouen route to Lower Normandy and site of the nunnery of Logium, last mentioned in 833. Continuing upstream there is Conihout, a hamlet six kilometres upstream from the Carolingian port of Jumièges Abbey; and to finish on the north bank, there is Sahurs, attested as a possession of the church of Bayeux in the eleventh century and a possible pre-Viking dependency of this church.
On the south bank, there is the same procession of Scandinavian place-names and the same relationship with the harbour sites. The series begins at the entrance of the bay with Honfleur, Crémanfleur and Fiquefleur, all three at the mouth of inshore streams. Next is Rislecliff at the mouth of the Risle, a port frequented in the pre-Viking period by ships from the abbeys of Pental and Saint-Germain-des-Prés. At Quillebeuf was a major portus, returned to Jumièges Abbey at the time of its restoration c. 942; further upstream were two ancient river crossings also part of the estates restored to Jumièges, the first at Vieux-Port, formerly Twuit-Port, on the old route from Lillebonne to Brionne, the second at Bliquetuit opposite Caudebec-en-Caux, on the route from Rouen to Brionne. Between the two was Brotonne, a replacement of the primitive place-name Arelaunum which indicates a royal Merovingian estate with its forest, palace and port on the Seine, last mentioned in its old form in a Jumièges diploma of 849. On the outskirts of Rouen and in the area of the town’s harbour dependencies, we find Couronne (comm. of Grande-Couronne and Petit-Couronne), whose pre-Viking name may be ignored, but where the archaeology and church dedications witness Merovingian antecedents; upstream from the town, one finds Elbeuf and Caudebec-lès-Elbeuf, site of the large ancient vicus of Uggade. In total, therefore, there are some twenty hamlets or localities, of which a large majority are ancient monastic ports, which received a new Scandinavian name in the tenth century.
To get a more exact idea of the importance of this wave of toponymic change, it remains to outline the list of ports surviving under their ancient or later Medieval names. Leaving aside three recently alleged cases of continuity which are far from secure,* we can only see in this series the names of Quenneport (comm. Val-de-la-Haye), generally identified as Quenzicoportus in a charter of 875; La Fontaine (comm. of Hénouville); Duclair, witnessed about 860 under the form Duro-clarum; Le Trait, site of a very old river crossing (trajectus); Aizier (canton of Quillebeuf) which derived from the ancient Aysiacus; and perhaps Jobles (comm. of Fatouville-Grestain), if its eleventh-century name Joninis can be equated with the Jovitinos./Jovitinis of a Fontenelle Abbey charter of 854. Even if no exact correlation is possible with these examples, the numbers are eloquent: only five or six names were maintained compared to about twenty cases of name changes.
How can this wave of toponymic change be interpreted? It is difficult to use the same argument as that advanced for estates renamed with a Scandinavian name in rural surroundings, for example in the Caux region; that interpretation ascribes changes to a minority Scandinavian presence with the local chiefs having received these places in freehold after 911. For the ports, the greater number of the centres concerned were taken into the public domain in the tenth and eleventh centuries. On the other hand, with the exception of a few dependencies like ‘the clearing of Belingethuyth’ or the small islands of the Seine exploited for their pasture like the Torholm islands (Oiseel) and Engohomme (by Martot), there is sparse evidence here for Scandinavian place-names based on an anthroponym; the vast majority of Scandinavian place-names in the Lower Seine are of a descriptive nature and allied to the Viking nomenclature of the Cotentin seaboard studied by René Lepelley. Here, as noticed by David Bates, the toponymic groupings of this type imply a mass peopling occurred in a relatively short time (Bates 1982, 18). In these circumstances the toponymic changes, with the renaming of old settlements, would be linked to a renewal of population density, that is, to the substitution of the indigenous population by Viking groups.
The circumstances of Viking settlement in the Lower Seine
From the foregoing, it follows that a good number of secondary settlements on the Seine, abandoned because of the transfer of their inhabitants to Rouen, were repopulated by groups of Scandinavian or Anglo-Scandinavian origin. It remains to be determined when this episode took place and under what conditions this installation of newcomers took place.
In all likelihood, the Vikings were unable to settle in these places before their evacuation by the Frankish population, and therefore not before the years 888–90, as it was only then that the transfer of inhabitants to Rouen took place. Neither can it have taken place in 890 because the Vikings, chased from Francia by King Odo, were then operating in t
he Cotentin. Moreover, it is precisely because they had been cleared from the Lower Seine region that it became a place of refuge for the clergy fleeing the Cotentin. When, at the end of 890 the Vikings were expelled from Brittany by Alan the Great, they re-entered the Seine but did not stop, proceeding directly to Noyon where they established their winter quarters. In 892 the last groups still operating in northern France and the Escaut region departed to raid in England. Only in 896 did a Viking fleet again enter the Seine estuary. Under the command of a chief to whom Frankish sources give the name Hundeus, this group was probably one of those mentioned by the Anglo-Saxon Chronicle, s.a. 896, in reporting Viking war bands returning to the Continent because they had been unable to settle in England and found themselves without resources.
Arriving in Francia, it seems that Hundeus had the intention of negotiating with the Carolingian authorities. He could not hope for any gesture from King Odo, hostile as always to any sort of arrangement with the Vikings. He therefore sought out Charles the Simple, whom Odo had driven out to Lotharingia; Charles proposed baptism for Hundeus, provoking the fury of Archbishop Fulk of Rheims. The acceptance of baptism by a Viking chief normally required some form of remuneration in land and silver but it is not known what was granted or promised to Hundeus. In any case, Charles was at that time hardly in a position to cede any land to the Viking chief in the west of the kingdom, then under Odo’s control. But, on the death of the latter in 898, Charles regained his crown, the entirety of his territories and the possibility of negotiating as he pleased with the Vikings. In these circumstances, it seems difficult to place the arrival of Rollo in Francia before 898. Supporting an early date is a passage by Adhemar of Chabannes, writing in the 1020s, calling to mind the shaky friendly terms between Rollo and Rannulf II of Poitiers, who died in 890 (Bauduin 1999, 12; Chavanon 1897). However, this late source is a feeble authority, especially as in the same sentence Adelmar describes the marriage of Ebalus Manzer, count of Poitiers with Rollo’s daughter Adèle, which is a flagrant error: Adèle’s husband was, more probably, William ‘Towhead’, son of Ebalus (Searle 1988, 54 and n. 74).
In his famous account of Rollo’s first entry into the Lower Seine, Dudo of Saint-Quentin reports that the Viking chief took his fleet as far as the Jumièges port of Saint-Vaast. There, according to Dudo, he received emissaries from the archbishop who asked him to spare the city of Rouen, which was only inhabited by a defenceless civil population and merchants without resources; Rollo promised not to attack (Lair 1865, 152–3). In spite of the serious reservations usually provoked by Dudo’s accounts (Yver 1969), in particular those concerning Rollo, this passage deserves the greatest attention. The place of the meeting is not known to have been invented and more than just the archbishop’s mediating role and the situation of Rouen’s population, all this tallies with the context at the end of the ninth century. On the other hand, it is clear that an old Jumièges tradition existed, remembering the ruin of the abbey in the ninth century and its reconstruction by William Longsword c. 942 (Lemarignier 1955). Today this source is lost, but fragments are preserved in William of Jumièges’ Gesta Normannorum ducorum (completed c. 1070: van Houts 1992, xxxviii); Dudo borrowed heavily from it, in particular to write the passage which he devotes to Jumièges and the work of William Longsword. On this basis, except in the details – the date of 876 assigned to this episode is evidently wrong – the accuracy of the Jumièges agreement is plausible.
It is a small step from this to the suggestion that Viking settlement in the Lower Seine was a direct consequence of this agreement. Property of the royal fisc, including these deserted ports, would have been ceded to the Vikings by the Carolingian authorities in exchange for a promise of non-aggression against the city of Rouen. This scenario of a ‘legal’ settlement seems the more likely. Indeed, there had been at least one precedent. In 877 Charles the Bald had surrendered the island of Bièce near Nantes, with the right to trade, to a group of Vikings who had escaped from the siege of Angers, in exchange for the promise that they would accept baptism. Thus, the Seine valley must have seen the establishment of a kind of cohabitation between the indigenous population, urban and peasant, and the Scandinavian or Anglo-Scandinavian groups settled in the harbour vici.
The situation was undoubtedly difficult for the religious communities of the Cotentin who had found refuge in the river area in 890: most of them left the region. The group occupying the port of Saint-Paul near Jumièges left with the body of St Clair to settle in a place in the Vexin which took the name Saint-Clair-sur-Epte; the monks of Saint-Marcouf left the port of Emendreville to reach Senlis, then Corbény; the Graville group fled to Conflans with the body of St Honorine, a translation which one source dates to the reign of Charles the Simple, therefore after 898. On the other hand, it seems that the Vikings gave sufficient guarantees to the people of Rouen for them to feel secure inside the city walls. As far as is known, the clergy from the church in Coutances, settled since 890 at Saint-Sauveur, did not leave the town, and it is highly probable that the count mentioned in a charter granted in 906 by Charles the Simple to his chancellor Ernustus had been the count of Rouen. Conforming to the Jumièges agreement, the city of the Lower Seine therefore remained a protected haven where there remained not only a large Frankish population, but also a religious and administrative framework. Knowing whether the mercantile suburb was part of the harbour vici ceded to the Vikings by the Jumièges agreement, is a complex problem, about which we can only spare a few words. It is not out of the question that some Vikings soon settled outside the walls, as happened at York. But, supposing this was the case, the occupation was not prominent: the Rouen suburbs have not preserved any place-name of Scandinavian origin. This fact is certainly not new. However, this perhaps gives insufficient prominence to Flodoard’s mention of the numerous departures that were provoked at Rouen among the Scandinavian population upon the entry into the town of the French after the death of William Longsword in 942.
* This chapter represents the summary of a study set out in more detail in a work in preparation on Rouen et les Vikings – De la Cite Carolingienne à la Ville Normande. One may find there all the references and arguments which it has not been possible to include in this present paper.
CHAPTER 19
San Vincenzo in the Making: The Discovery of an Early Medieval Production Site on the East Bank of the Volturno
Matthew Moran
Introduction
Much is now known of the archaeology of inland production and trading sites in Early Medieval northern Europe, and their economic context. The same may not be said of Italy, however, where our knowledge of commodity production and exchange is limited to sites on the seaboard, and exceptional inland contexts such as Rome. Away from the coasts, in city and countryside alike, commerce and patronage were based above all on land ownership and largesse. Thus the making of San Vincenzo al Volturno coincided with a period, c. 800–50, in which the monastery received from private donors gifts of land in all parts of Beneventan territory. This paper considers how the monastery translated its immense agricultural revenues into prestige artefacts and buildings. Particular attention is given to newly excavated evidence of a production site established c. 820 alongside the monastery, on the opposite bank of the River Volturno.
FIGURE 19.1. The location of San Vincenzo (a) and a plan of the site (b).
San Vincenzo al Volturno
San Vincenzo al Volturno stands high in the upper Volturno valley, in the modern region of Molise, approximately 170km south-east of Rome, and about 30km north-east of Monte Cassino (Fig. 19.1a). The site is approximately 530m above sea level, in the lower Mainarde, the foothills of the Apennines (Fig. 19.1b). San Vincenzo is almost equidistant from the east and west coasts of the Italian peninsula – or as far inland as it is possible to get. British-led excavations at San Vincenzo al Volturno began in 1980, with two principle objectives – first, to identify the location and aspect of the Early Medieval monastery; and, second, to examine
the monastery’s relationship to settlement in the landscape (Hodges 1993, 6).
Prior to c. 780, and despite the institution’s growing political and cultural significance (Wickham 1995, 138–52), San Vincenzo was a materially unprepossessing place. The monastery’s founders evidently restored parts of a fifth-century villa using pre-existing building stone and tiles, mortared with clay, but otherwise, there is practically no evidence of material culture production in the monastery (Hodges 1993, 128–34). By contrast, a structured reconfiguration occurred in the period c. 790–840, which was made possible by the presence and production of people skilled in tile-making, bronze-casting, and iron-and glass-working (Hodges and Mitchell 1996; Mitchell 1994b; Mitchell and Hansen 2001). The nucleus of this production was the monumental abbey-church, San Vincenzo Maggiore, beneath which a sequence of temporary tile, bronze and glass workshops was excavated (Mitchell 1996; Francis and Moran 1997).