by Tim Pestell
With these reservations in mind, some comment may be offered. The importation of both metal, such as iron ore, and pottery at Tjitsma indicates the site’s participation in wider trade networks, despite the evidence for only small-scale craft production within the terp. Its location may have much to do with this. Tjitsma is located close to the sea, and to the important trade routes from Dorestad to England, north Germany and Scandinavia, and to the river systems heading inland. Interestingly, despite this, Wijnaldum appears never to have grown into a major trade centre. Within the wider context of the terpen, the discovery of metal objects appear to say something about the importance of some sites despite a lack of excavation.
Many gold finds are known from sixth-and seventh-century Friesland and among the metal-detector finds from terpen, some appear to be richer in finds than others. For example, Dongjum terp is quite rich, while a large gold deposit was found in Wiewerd. Wijnaldum similarly appears to be among the richer terpen, although not on the same scale as Scandinavian production centres like Birke or Helgo. Nevertheless, its imported goods and rich metalwork assemblage indicates that it constituted a small production centre engaged in trade and presumably controlled by a local chief. The principal difficulty remains the need for qualitative data from other terpen with which to compare Wijnaldum – and this provides the challenge for future archaeological investigation.
Acknowledgements
I would like to thank the following people and institutions: the Laboratory of Conservation and Material Sciences of the Groningen Institute of Archaeology (GIA), University of Groningen (RUG); Jurjen Bos and Danny Gerrets for their cooperation; Justine Bayley and Nigel Meeks for helping me with the analyses of several finds; English Heritage Ancient Monuments Laboratory and the Department of Scientific Research of the British Museum for the use of their facilities; and the Netherlands Organisation for Scientific Research (NWO) for funding my internship at English Heritage.
CHAPTER 18
The Fate of the Ports of the Lower Seine Valley at the End of the Ninth Century
Jacques Le Maho (translated by Joy Southam)
This chapter investigates the consequences of the Scandinavian raids on the ports of the Lower Seine valley in the second half of the ninth century. It examines: the impact of these raids during the period from the first Viking incursion in 841 to the lifting of the siege of Paris in 886; the consequences of transforming the city of Rouen into a fortified town in 889–90; the population movements that textual and toponymic sources suggest in the Lower Seine area between the end of the 880s and the end of the ninth century; and the circumstances of the establishment of the first Viking colonies.
The peopling of the Lower Seine before the end of the 880s
In Rouen, a distinction needs to be made between the area surrounded by city walls, and the suburban area. The seat of Notre Dame cathedral and its chapter, of several monasteries, of the count’s palace and doubtless many residences of the local leading citizens, Rouen was a centre of administrative and religious power. The intramural area, which covered close on 18 hectares, was probably quite sparsely populated. Around the houses of the clergy and the monastic cloisters extended courtyards, gardens and vast orchards, and the city almost certainly contained a number of open spaces. The likely presence of quite a number of private courtyards, with their buildings scattered randomly, only reinforces the impression of a semi-rural landscape, similar to the appearance of Winchester and London during the first centuries of the Anglo-Saxon period.
Outside the walls, the settlement pattern differed. Though numerous spaces were not built up, and vast cemeteries extended around the funerary churches, the artisan quarters were located in this suburbium. Different portus assured the provisioning of the town. We may recall two well-known texts from the eighth and ninth centuries, the Life of St Winnebald where a market ‘near to the city of Rouen’ is mentioned, and the account in Nithard’s History of the Sons of Louis the Pious, of the arrival in 841 of a convoy of twenty-eight merchant ships, to a harbour site on the left bank of the Seine and therefore opposite the city (Scholz and Rogers 1972, 147). The main population of craftsmen and tradesmen therefore resided in an area outside the walls, doubtless scattered in quite a large district around the city. Finally, on the eve of the first Viking raid, a considerable part of the revenues of the portus of Rouen, exactly half according to a confirmation charter of 876, went to the Rouen abbey of Saint-Ouen, the remainder of the taxes probably entering fiscal coffers (Le Maho 1995); this point is of importance as we will see later.
If the presence of an important royal mint in Rouen tends to focus the attention of Carolingian economic historians on this town, it should not be forgotten that the Lower Seine region sheltered numerous other ports up to the end of the ninth century. Most of these were monastic properties. The abbeys situated in immediate proximity to the river, such as Jumièges, Fontenelle (Saint-Wandrille-Rançon), and Logium (Caudebec-en-Caux) on the north bank, Pentale (Saint-Samson-de-la-Roque) and Pennante (Saint-Pierre-du-val) on the south bank, made use of their own ports on the Seine, to which were occasionally added annexed sites like those of Jumièges at Duclair, at Vieux-Port and at Quillebeuf. The abbeys a little further from the Seine also had access to the river. It is quite probable that the monks of Montivilliers, for example, had berths for their ships at Harfleur in the pre-Viking period. Other vici harbour installations belonged to establishments such as Sanvic at the entrance to the bay of the Seine, probable site of the portus of the abbey of Saint-Denis, through which passed a cargo of lead from Quentovic in 810; Moriniacus, an unidentified river port local to Rouen, which a text of 860 mentions as the property of Corbion Abbey (Saint-Lhomer) in the Perche; or even Pont-de-l’Arche above Rouen, perhaps a possession of the monastery of Saint-Vigor de Bayeux.
Fishing must have represented an important activity to these monastic ports, but they also ensured the passage of men and goods, whether over short distances – it seems that most of these crossing points had been under the control of the abbeys – or, for more important journeys. It was by boat that the principal monasteries of the Seine like Fontenelle and Jumièges maintained communication with their estates of the Parisis, Cotentin, Bessin or Boulonnais regions. The involvement of these abbeys in Channel trade is also well known. Documentary sources, however, give us little information on the importance of settlements associated with these ports. Most often one has to be content with vague hints, like the prosperity which trade with England brought to the Jumièges region in the lifetime of St Philibert (eighth to ninth century). We have only two indications of figures in our sources: the taking of sixty-eight hostages by the Vikings from the port of the abbey of Saint-Denis (without doubt Sanvic); and the mention of the twenty-eight smallholdings in the port of Fontenelle belonging to the monastic estate. These references suggests the existence of settlements, extensive in the second case, indicating that the port ought to be distinct from the monastic vicus closer to the abbey, where metals, leathers and textiles delivered in large quantities to the monastery were processed. The impression which emerges from this brief outline of the Lower Seine landscape before the first Viking raids is that of an active and populated countryside, but in which economic activities were scattered in a nebula of harbour-sites or of craft and merchant vici, without a real preponderance appearing around Rouen.
In spite of repeated incursions by Scandinavian war bands from 841, this system held together without much modification until the end of the 880s. The abbeys were, however, nearly all evacuated before this date. The monks from Saint-Ouen, Rouen, found refuge at Gasny (Eure) towards the beginning of the 860s, then, after the Viking raid of 876, to Condé-sur-Aisne (Aisne), near Soissons. The move must have been followed by other Rouen communities, like those caring for the relics of St Mellon and St Filleul, while the body of St Austreberthe, foundress of the monastery of Pavilly, was taken to Saint-Valéry-sur-Somme and then to Ponthieu. The monks of Fontenelle left as early as 858 on a long exo
dus which took them as far as Montreuil-sur-Mer and to Boulogne-sur-Mer. As for those at Jumièges, after a number of episodes, their wandering ended at Haspres (Nord) in the Cambrai area. At the start of the 880s it is likely that not a single monastery operated in the Lower Seine area.
Despite this, the activity of the ports had not ceased. From their coastal refuges of Boulogne and Ponthieu, the monks of Fontenelle had kept in contact with their lands on the Seine, and a text from the abbey of Centule suggests that the abbot of Jumièges, who had withdrawn to Saint-Riquier in the 860s, was doing the same. Even at Rouen, the royal mint continued to strike coinage in abundance after 864, demonstrating the maintenance of economic activity and survival of commercial links between the city of the Lower Seine and Quen-tovic. Saint-Ouen’s possession of half the revenues from the town portus was confirmed in 876, and there is no reason to think that these rights fell into disuse. In 885, the Vikings arriving at Rouen from the interior of the country found enough merchant boats on site to get their whole army across the river. From all this, it seems that a large part of the active population had stayed in the area, in other words that the vici of the Seine were still inhabited. Even in Rouen, there is no indication of a withdrawal by the inhabitants from the city’s suburbs. Like the example of Nantes, it presents all the characteristics of an ‘open town’ and it is apparently into the almost deserted Roman castrum of Rouen that the Vikings entered on at least two occasions, first in 851, and again in 885. Revealingly, Archbishop Venilon had a refuge fitted out at Andely in the Vexin from the beginning of the 860s while in 876 Archbishop John finally had to leave the province and retreated to his residence at Braine in the Soissons area, where he was joined by his canons. Nothing indicates that the archbishop returned to Rouen before the end of the 880s.
Rouen: a city of refuge from 888 to 890
During the winter of 889–890, the Vikings who had just been repelled from the Parisian region by Odo, made their way towards the Cotentin and laid siege to the town of Saint-Lô. A large number of inhabitants took refuge in this stronghold, including Lista, bishop of Coutances. The town was taken and all the population was killed, including the bishop. The Vikings continued their progress into the interior of the country without meeting much resistance. They were only driven away in the autumn of 890, by a coalition of Breton forces commanded by Alan the Great. Meanwhile, the Viking attacks had provoked a massive exodus of clerics and religious communities from the Cotentin and from the extreme west of the Bessin area.
In dealing with these multiple movements, thirteen groups can be identified who moved to the east of the province with their relics. Several of them settled in Rouen itself. Bishop Ragenard, Lista’s successor to Coutances, settled with his canons in the city’s church of Saint-Saveur. The probable site of an ancient monastery, this church and its adjoining lands, was granted to them by King Odo (the alleged gift by Rollo to Bishop Thierry in 913 is, to all appearances, only a confirmation). Soon rechristened with the title of Saint-Lô, Saint-Sa-veur’s church was to remain the see for the bishops of Coutances until the beginning of the eleventh century. The monks of the abbey of Nantus (Saint-Marcouf, Manche) settled opposite the city at the port of Emendreville, while another group occupied a chapel near the high south defensive wall with the relics of St Candide, an obscure individual venerated at Picauville (Manche). We should also mention the probable journey to Rouen of a fourth group with the relics of St Leon of Coutances, a fifth with the body of St Germain the Scot and a sixth with that of St Jean originally from Deux-Jumeaux (Calvados) in the diocese of Bayeux. All left the Cotentin or the western borders of the Bessin, probably at the end of 889 or at the beginning of 890 (Le Maho 1999).
The fact that these monks in exile chose the city of Rouen to fall back to, with the obvious intention of settling there, shows the situation in the town had changed radically. In fact, archaeology bears witness to major transformations in the city district. Part of the high north defensive wall of the castrum was reconstructed at the end of the ninth century, as was the Beauvais gate at the northern entrance to the ancient cardo maximus. In the interior of the city itself, the road system was the subject of a complete reconstruction. Numerous north-south axes, inherited from the Gallo-Roman system were abandoned and new roads developed, spreading out to east and west at a right angle to a central north-south road (Gauthiez 1998, 81–3). At the same time, archaeological finds witness a rapid and considerable increase in the density of settlement inside the walled city. Close to the cathedral, excavations from 1980–90 brought to light remains of two small blocks of dwellings established at the end of the ninth century on the edge of the Rue Saint-Romain and the Rue du Change, on ground previously occupied by the canonical cloister and buildings of the archbishop’s palace. Their construction coincides with the establishment of the new road system and it was probably the same for numerous other settlement sites located inside the city, but which have so far only been revealed as courtyard floors and clusters of domestic pits; the terminus post quem of these are invariably the end of the ninth or beginning of the tenth centuries. The houses brought to light next to the cathedral were regular in plan, constituting enclosures in strips perpendicular to the road with a constant width of six metres.
Thus, when the religious groups chased from the Cotentin came to settle in Rouen, it seems that the town’s fortifications were repaired and that the intramural zone was the subject of an important urban redevelopment, accompanied by a general restructuring of the land register and a parcelling out of the land plots. This raises a series of questions, one of them being to whom the initiative for this operation can be attributed. The archbishop might be considered first, on the assumption that he had returned to Rouen in 889–90, which is not certain; on the other hand it is clear that his long absence had enabled the extension of secular power in a large part of the town to his cost. Originally, large areas of the city belonged to the monasteries and to notable families. But, when it had become obvious that the religious communities in their distant refuges were no longer in a position to administer their estates in Rouen, these monastic properties had been united into the charge of the royal fisc. This is known for the monastery of Saint-Saveur, in the king’s hands in 889–90, and it is very probably the case for Saint-Amand, which was in Rollo’s hands after 911. As for the properties of the great Rouen families, they were also destined to rejoin the public domain once the city was declared loca deserta; one can assume that the flight of the urban elite, because of the Viking danger, had considerably increased the number of properties in escheat. At the end of the ninth century, therefore, the king had almost complete control over the whole city, reserving the estate of the archbishop and chapter. It follows that the reorganisation of this urban space resulted from an initiative by King Odo (888–98).
The election of Odo by the elders of the kingdom had been the political reward for his brilliant victory over the Vikings at the siege of Paris (885–6). The first major success achieved over the Scandinavian invaders, this victory restored confidence to the Frankish authorities and encouraged the establishment of a complete defensive infrastructure in Francia. Castella were built just about everywhere by counts and abbots, and a number of cities were transformed into refuge-towns. The case of Rouen is therefore not isolated. However, in comparison to Chartres, Angers and Rheims, for example, the city was almost empty of inhabitants when it was subject to this urban reconstruction. The question is therefore to determine the origins of the civilian population which then came to repopulate the city, established in a space designed and refitted for the circumstances.
We might first consider the refugees coming from the surrounding countryside. But contrary to this hypothesis, toponymic data, as early as the documentary and archaeological evidence, indicates a great continuity in the occupation of peasant villages in the Seine valley throughout the period under consideration. On the other hand, there are numerous signs of a strong concentration of artisans and merchants in the c
ity. In a passage doubtless inspired by a tradition from Jumièges – to which we will return – Dudo of Saint-Quentin (writing 1015X1026) alluded to merchant refugees in Rouen. A famous document from the abbey of Saint-Denis, the pseudo-gift of Dagobert (621–37) but dated variously to the end of the ninth or beginning of the tenth century, also mentions the ‘merchants of Rouen’. Finally, if the archaeological data is quite sparse on the subject, it remains significant that the only two areas of intramural settlement dating from the beginning of the tenth century to have been the subject of extensive excavations, the cathedral quarter and the Lycée Camille-Saint-Saêns, have yielded, in the former, traces of metal-working activity and in the latter, waste from bone-working. It is a small step to imagine that at the end of the ninth century refugees made up the core of the large population residing in the city into the middle ages, whose professions came to provide several street names in this part of the city, including those of exchange, ironworkers and hosiers.
These observations lead us to suppose that at least a part of the population within the city at the end of the ninth century came from the ancient harbour and mercantile districts of the Rouen suburbs. Their transfer into the walled zone by Odo might be compared with the operation carried out by Alfred the Great for the transfer of the merchant suburb of Lundenwic into the city of London. But the transferred population seems much too large to have come only from the immediate area. Since the departure of the monks, the ports and monastic vici of the Lower Seine had maintained their activity to a creditable level, but there came a point at which the religious communities in exile were no longer in a state to administer them. One after another, most of these vici ended by falling into the royal fisc, with the exception, perhaps, of some riverside settlements still in the possession of powerful lay abbots who themselves owed their benefices to royal favour. We may therefore suggest that at the end of the 880s, the vast majority of the harbour vici of the Seine were in the hands of the king. Consequently, there are good reasons for thinking that the inhabitants of these vici, craftsmen and merchants, constituted the largest part of the population settled in the city of Rouen. This parallels other well-known cases, for instance the transferral to Montreuil-sur-Mer of the portus of the Lower Canche valley, and of the inhabitants of Hamwic and the vici of the Southampton region, into the city of Winchester. By relocating the main merchant population of the countryside in the same urban site, the organisers of the operation clearly aimed to control these human resources better and to draw from them the maximum profit. We must not, therefore, misjudge the real character of these population movements; far from being spontaneous, they were planned, organised and even accomplished under constraints. Indeed, we may recall a famous precedent, the transfer to Hedeby in 808 by the Danish king Godfred, of the whole population of the port of Reric, identified as Groß Strömkendorf, in the bay of Wismar at Mecklenburg (Tummuscheit, this vol.). Here, the deportees – they cannot be called anything else – had to complete a crossing of almost 130km.