Comparative Studies on Early Medieval Harbours

Dr. Sven Kalmring

The harbours of Dorestad (NL) and Birka (S)

Dorestad
The Late Merovingian and Carolingian trading centre of Dorestad (mid-7th to mid-9th century AD), located at the junction of the rivers Kromme Rijn (Crooked Rhine) and Lek in The Netherlands, functioned as a kind of hub for Frankish trade. From here, trade contacts extended out across the entire Rhine and Meuse area, to England, to Northern Europe and into the Baltic region.

During excavations carried out between 1967 and 1977, efforts were concentrated on a contiguous part of the actual habitation area, comprising the northern section of the settlement and the cemetery de Heul belonging to it; the former bank of the Kromme Rijn was similarly investigated. Here, to the right of the Hoogstraat, an old track on the embankments of the Rhine, regular rows of piles were uncovered which had formed part of dike-like causeways; wooden-earthen structures leading into berths for the ships.

The siting of Dorestad’s harbour area at Hoogstraat seems initially to have involved a particularly favourable part of the Rhine river bank, as the harbour was located on an inner loop of the river and the bank was therefore not exposed to any erosion processes. The bank here commanded a flattish slope with a sandy shore on which the ships could land. Gradually, however, the river began to meander and its bed shifted continuously towards the east such that a wet and low-lying area of land, almost a shoal, was formed.

In order to maintain the connection between the bank and the berths, vital to the settlement, causeways were constructed involving great expenditure of labour and timber. In the construction process, a casement was formed using sharpened piles driven deep into the ground linked by wattlework cladding; this was then filled with soil. The soil needed for the fill came from the ditches which could always be observed between smaller groups of piles.

However, the extension of the wood-earth constructions took place gradually, in time with the continuing relocation of the Rhine, and involved numerous successive construction phases which mostly served to extend the length of the causeways. This expansion of the overall complex of causeways took place over a period of around a century and a half. In the early 9th century, the enlargement of the structures ceased. At this point in time they extended out over a total length of c. 200 m.

According to new interpretations by the excavators W.A. van Es and W.J.H. Verwers, the extensive works on the former area of the riverbed did not, however, serve primarily as a way of further exploiting the land as a harbour, but were rather a process of land reclamation for settlement purposes. Timber structures within the causeways, which previously were seen as an indication of planked surfaces, i.e. boardwalks, are now interpreted as pile-foundations from constructions with a granary-like appearance. Against the background of the excavations in Schleswig the new interpretation of these otherwise well-published findings will be critically reviewed within the framework of an article.

The results of the study were presented at the conference "Dorestad in an International Framework" 2009 in Leiden, and at the 29th Tværfaglige Vikingesymposium 2010 at the Viking Museum Haithabu. In addition, the results were presented in the following articles:

S. Kalmring, Dorestad Hoogstraat from a Hedeby/Schleswig point of view. In:

• A. Willemsen (ed.), Dorestad in an international Framework. New Research on Centres of Trade and Coinage in Carolingian Times [Symposium Leiden 2009] (Turnhout 2010) 68-81.
• S. Kalmring, Dorestad Hoogstraat. Ein Diskurs gegen das Verschwinden des Hafens des „vicus famosus“. 29. Tværfaglige Vikingesymposium Schleswig (Højbjerg 2012) 22-49.

Birka
The Viking Age trading centre of Birka (8th to final quarter of 10th century AD) is located on the island of Björkö in Lake Mälar, around 30 km west of Stockholm. Unlike the situation today, the Mälar area in Viking times was still an extensive marine bay on the Baltic. It was first in the 10th century, through the post-glacial rebound, that it gradually became separated from the sea. Situated at the junction of the two most important sailing routes in Central Sweden, the early town was located in a strategically very favourable position with regard to long-distance trade.

On the one hand, an over-land transport route, running from the interior in the direction of the Baltic along an east-west axis, ran through here. Relative to this route, Birka constituted an important hinterland and stockyard for commodities such as iron or wood. These were transported over land to Birka from Central Sweden during the winter on sledges and then, during the summer months, they were shipped onwards. The second over-land transport axis was the so-called Fyrsleden. This ran from the Baltic in the south, across the sound at Södertälje in Lake Mälar and continued via Birka in the direction of Sigtuna, Gamla Uppsala and Vendel, north in the Mälar valley.

Through excavations carried out by B. Ambrosiani and B. Arrhenius in the years 1969-1971, down through deposits arising during the continuing regression and dramatically falling water levels, a jetty was documented in the harbour area of Birka which is now connected with the mainland. The main aims of the excavations in the area with the svarta jorden (i.e. black earth) were location of the Viking Age coastline and an investigation of the potential harbour area. One of two cobbled pavements was excavated and exposed as the landward bi-partite foundation of a landing stage. It had been constructed in the damp shore area in the vicinity of a sand bank, which had formed around a bank reinforcement constructed of round-wood and which had presumably functioned as reinforcement for the jetty. On the basis of stratified finds recovered during the course of the excavation, the recorded stratigraphy is dated to a period extending from the middle of the 9th century until Birka’s final phase around AD 970. The jetty (‘final jetty’) itself is assigned to within the period AD 930-950.

Beneath the landing stage from the 10th century there were indications of an older, similarly constructed structure, embedded in a layer dating from the 9th century (‘lower jetty). During the period 1990-1995, a further excavation was carried out in an area above that of the 1969-1971 excavations. In the excavation trench from 1990, a layer of stones up till 0.8 m in size protruded from the eastern wall; this was interpreted as the face of a further jetty (‘earliest jetty’). It was located at a height of around 6.1-6.3 m a. s. l.

Other harbour basins are considered in the discussion: Kugghamn, Korshamn and Salviksgropen. Further, as yet un-investigated, stone foundations of two jetties are located away from the trading port, below the garrison. In the southern part of the Björkö island, near Charlottenlund, there is a boat house of presumed naust type. Furthermore, since 1970, diving campaigns have been carried out at regular intervals in the area off the svarta jorden; these were intended to explore the underwater part of the harbour basin. Various cores arising from an auger survey carried out in 1990-1991 also provide information on the structure of the harbour basin. And in 2007, a still unpublished geophysical survey was carried out on land which contains further important evidence.

Within the framework of an extensive study, and in close co-operation with our Swedish colleagues, all the harbour-related information derived from excavations, diving campaigns and geophysical surveys is going to be collated and then illuminated on the basis of the insights gained at Haithabu.