Om en vestjysk stald

Abstract

A West-Jutland Cattle-house

In the course of investigations of Iron Age settlements in the Esbjerg area there have on three occasions been found house sites from the Early Roman Iron Age which are distinguished from all previously known examples by having paved stock quarters.

It was as early as 1932, when the investigations of the Jutland settlements were at their height, that the first house came to light at Kjærsing in Brøndum parish, Ribe county ²). It showed a well-preserved portion of paving with a clearly sunk gutter in the middle. This novel feature of the Kjærsing house gained renewed interest when still another house with paved stock quarters was revealed in 1945 at a settlement site in Vognsbøl, in Jerne parish just north of Esbjerg ³). Later, in 1952, the third paved stabling was found, this time at Boldesager in Jerne parish.

The isolated occurrence of this type of animal quarters in these three houses, only a few miles apart, suggests that this paved stabling of the Iron Age is a phenomenon localised in this district, though even so it cannot be regarded as the norm, as three other house sites investigated in the same district had no paving.

The two investigations first named were carried out under the auspices of the National Museum, whereas the last named, the Boldesager house, was excavated by Esbjerg Museum and the Esbjerg Evening College. It was found in the course of laying sewage pipes to serve new streets - which results in the diagram of the Iron Age stock quarters, which was the result of the investigation (fig. 1), being disfigured by two ditches cutting through the excavated area.

It was nevertheless possible to obtain a fair survey of about 14 metres of the house, which was oriented approximately east-west; of this about 12 metres must have comprised the animals' quarters, while the dwelling area, which undoubtedly lay - against normal custom - in the east end of the house, was unfortunately on a piece of land which could not be excavated.

Two rows of free-standing, roof-bearing posts had stood within the house, the carbonised remains of several of which were still to be seen in the post holes, showing that the house was destroyed by fire. The walls had been constructed of small vertical posts set in the ground, presumably supporting a wattle and daub structure, though this could not be demonstrated. The house was about five metres broad.

Within this frame a quite clear picture of an Iron Age farmer's stock quarters appeared. The paving uncovered outside the north wall should perhaps be understood as the paving before the entrance to these quarters, even though the resemblance to such a paving was only slight (fig. 2).

In the best preserved portion of the paving the construction of the stock quarters can be clearly seen (fig. 3). A depression, about 20 cms. deep and bordered by stones standing on edge, runs through the centre of the house, and could also be identified with certainty at the very disturbed eastern termination of the paving. On both sides of this gutter the stalls for the animals had stood between the free-standing posts and the outer wall. In several places their presence could be demonstrated, partly by wear in the paving and partly as dark tunnels dug in the gravel subsoil. There were, it seems, 10-12 stalls in the stock quarters, each with room for two animals. Such stall partitions have been identified at various other sites, including those of Skørbæk Hede ⁴) and Gørding Hede ⁵).

It is possible that the westermost end of the house was completely divided off from the stock quarters by a party wall, as a foundation trench there extends completely across the house from wall to wall. This circumstance, fortified by the discovery of food remains (mussel shells) and a small amount of pottery, suggests that the west end was a humble dwelling - the stable-boy's room, if we will.

The real dwelling area of the house must be sought in the east end. Along the eastern termination of the stable paving a quantity of carbonised wood­work was identified, presumably fallen down or overturned at the time of the fire, and probably forming the remains of a party wall between the living quarters and the stock area. About one metre east of this party wall a typical Iron Age hearth was found, with a thick clay layer above a stone setting. This presumably belonged to the part of the house which, as mentioned above, extended into an inaccessible area.

Very little pottery was discovered, as could only be expected, as pottery is at home in the living quarters. However, in addition to the occurrence already mentioned in the west end, a number of scattered fragments were found, sufficient to date the building of the house with certainty to the Early Roman Iron Age. One discovery is worthy of remark, that of the broken-off horn from a hollow earthernware "fire-dog", of a type which is otherwise only known from the Vognsbøl site only about one kilometer from Boldesager ⁶). In this district several of these broken-off horns have been found, including some in the so-called "rubbish pits" and in a presumed cremation-patch burial. These discoveries of isolated horns may well give rise to speculation, as one of the Vognsbøl dogs was found in a burnt-out house, apparently in use with both horns removed.

Thus three houses have been found within a very limited area, distinguished from the normal run by having paved stock quarters. There were certain differences in the three layouts, both of dwelling and of stock quarters. These differences are not necessarily of any great significance and are only recorded here. for purposes of comparison.

The Vognsbøl house (fig. 4) had walls of earth, living quarters in the west and stabling in the east. The paving was here bounded by the roof-posts, and the gutter was merely a trench dug in the earth without a stone lining. Worthy of note was the fine paving at the entrance to the stock quarters on the northern side of the house.

The Kjærsing house (fig. 5) had no demonstrable interior posts. Traces of posts along the outer walls of the house suggest that these were of wattle and daub, as there is also reason to believe in the case of the Boldesager house. A detail of the paving which was not found in the other two stock quarters was a broad and extremely regular stone paving which appeared to separate the stock area behind the gutter from the dwelling quarters. Here too the stabling was in the east end.

The material available is not sufficient to enable the paved stabling to be given a definite place within the scheme of Iron Age discoveries. At the moment the only point which can be made is that the type has not been found elsewhere Concerning its extension in the Esbjerg area it can as yet only be recorded that of six house sites from the Roman Period three possessed stock paving. More extensive investigations in this area and in South Jutland might well alter this picture, and show that the Esbjerg houses are perhaps no local phenomenon but rather a northernmost extension of trends in Iron Age building customs which came from the south.

Niels Thomsen