Alrum – Brandtomter i en vestjysk byhøj fra ældre jernalder


  • Jørgen Lund
  • Poul Nissen


Alrum, brandtomt, ældre jernalder


Burnt houses at an Early Iron Age tell site in Western Jutland

The Alrum settlement is renowned in particular for producing one of the largest prehistoric finds of charred grain and seeds ever discovered in Denmark.

The site was excavated in 1939 under the direction of Gudmund Hatt, but it was Hans Helbæk who carried out a detailed analysis of the plant remains. The latter were subjected to re-examination in 1994, whereas the extensive finds assemblage, stored at Ringkøbing Museum, has only now been fully investigated and analysed. The reason for this is that the excavation records, thought for many years to have been lost, turned up by chance at the National Museum of Denmark in 2000.

The Alrum site is located on a slight elevation, about 1 km from Stadil Fjord and 10 km north of the town of Ringkøbing (fig. 1).

The settlement
The excavation trench exposed an area of about 300 m2, within which there were sequences of six to seven house sites lying one on top of the other, resulting in cultural deposits with a vertical stratigraphy of 1.5 m, in other words a tell site (fig. 2). Two of the houses (house I and house II) had been destroyed by fire and had been abandoned in such great haste that everything remained within the burnt-out remains of the buildings. House II was the better preserved of the two, containing building timbers, c. 50 pottery vessels, straw ropes, some stone tools, a ball of wool etc. The house was 14.5 m long and 4.5 m wide (c. 60 m2), with living quarters at the western end and a presumed byre to the east. Relative to contemporary houses in Eastern Jutland, those in Western Jutland were small. The roof was borne by five pairs of posts arranged along the length of the house and was probably comprised of heather turf. The post-built walls had an inner cladding of thick oak planks, whereas the outer surface is presumed to have been covered with a layer of straw or grass. The living quarters were fitted out with a clay bench or platform at the gable, an ornamented hearth in the middle and, between the two, a stone mortar set firmly into the clay floor (fig. 3). No traces were seen in the byre of the usual stall dividers, so perhaps the house had not been fully completed when the fire broke out! Most of the pottery lay close to the clay bench, together with several bodies of untempered clay; these weighed c. 9 kg. Up against the north wall there were two impressive solid andirons, 27-28 cm in height and weighing more than 3 kg (figs. 4, 5, 6, 7 and 8). The pottery dates the house to the late Pre-Roman Iron Age.

Beneath house II lay the successive remains of six to seven other houses. The pottery reveals that this small village was founded around 500 BC, whereas the latest examples are from the century around the birth of Christ (figs. 6 and 13). Only parts of house I could be excavated, but here too a great deal of pottery was encountered, together with a few stone artefacts (figs. 9 and 11).

Building timber
Virtually all the woodwork in the burnt houses was of oak, supplemented by a little willow and alder which are well suited to making the wattle of the walls. In each house there was a large number of roof and wall postholes, with the charred post ends still in situ; along the walls lay large pieces of so-called wattle panels. As a consequence, it was possible to measure the dimensions of the timbers. Charring leads to a reduction in size of the timber, but by how much? Information received from the Danish Institute of Fire and Security Technology states that, as a rule of thumb, there is a reduction of 0.5-0.6 mm for every minute the fire burns. Figure 10 gives the timber dimensions alongside a column showing measurements after 20 minutes of burning, to which 1 cm has been added. In spite of the latter, the timber dimensions were still markedly less than those of unburnt posts seen at for example Feddersen Wierde in the North German salt marshes. As oak is totally dominant as the building timber, this begs the question as to where it was obtained? A pollen diagram from a site located 4-5 km from Alrum shows that the landscape was open and unlikely to have had large areas of oak woodland. One possibility is that the oak wood was obtained from Eastern Jutland, perhaps being exchanged for fish and other marine resources?

Agriculture and fishing
The large quantities of charred grain and seeds recovered from the site constitute an excellent basis on which to gain a detailed insight into the subsistence. The most important cereals were barley and oats, accompanied by a little wheat, flax and gold of pleasure. In addition to these, seeds had been gathered from a range of weedy species, with corn spurrey, goosefoot, and persicaria being the commonest (fig. 14). These weeds show that the arable fields were sandy and only lightly manured and this conclusion is supported by the size of the cereal grains which is also very modest. It seems likely that the low-lying fields were flooded with salt water from time to time, but barley, flax and gold of pleasure are all salt tolerant.

In historical times seeds of the above weed species were used in bread, porridge and gruel by farmers living on the Jutland heath. Tubers of false oat grass were also found at Alrum; these are rich in starch and therefore represent a good food supplement. The heaps of crop plant remains can be classified as threshed and unthreshed (fig. 15). This can perhaps give an indication of the time of year at which the fire took place; it was most probably in the autumn. On the other hand, the bone material from the site is very limited due to the well-drained acid sandy soil. Mention can, however, be made of a perforated ox astragalus (fig. 11a-b). Even so, it can safely be presumed that the many good grazing areas were extensively exploited.

On the basis of the site’s location and finds of stone net sinkers, it seems justified to refer to Alrum’s inhabitants as fisher-farmers.

Settlement and landscape
Today, the Jutland west coast has a harsh climate with sand drift and storms as significant factors in the lives of the inhabitants. But this was not always the case and in the Early Iron Age the situation must have been quite different: Sand drift was less extensive, the coastline had a different appearance and the sea level fluctuated, as can be seen for example at Højbjerg just south of Ringkøbing Fjord and in several other locations (fig. 1). A rise in sea level of just 0.5 m would reduce the area of shore meadow considerably (fig. 16). The woodland picture was also different.

The most important indicator of this very different landscape and environment is the sustained habitation which characterises many settlements, and is exemplified by Alrum with more than 500 years of activity at the same location, and even a further couple of centuries close by, as suggested by recent aerial photographs. People lived at Nørre Fjand for 300-400 years and Klegod, now located directly on the present-day coastline, was probably occupied for at least a century. Such extended occupation of the same site must also be presumed to have resulted in social and family-related changes.

There was of course some sand drift in the Early Iron Age. This is apparent from sand layers between the individual house phases and on the arable fields. However, it was apparently not so extensive that it prompted people to move; the sand layers are modest in their thickness. A good example of the stubbornness of these Iron Age people is seen at the small village of Klegod where the inhabitants ploughed through a layer of sandy soil of no less than 40 cm in thickness.

The course of the coastline must also have been quite different back then as remains of Iron Age settlements are revealed now and then by today’s fierce winter storms which can cut deep into the sand dunes. Klegod, which dates from c. 500 BC, is just such a locality and provides secure proof that the coast must have lain a good way out to the west at the time, perhaps as much as several kilometres.

In this dynamic and changeable landscape, the fisher-farmers of the Early Iron Age managed to maintain their existence over many generations and they were perhaps not as isolated as one could easily imagine. However, one main question remains: What led these people to settle in these near-coastal areas? The numerous Iron Age sites show that many families must have been involved. Was it marine resources or the good grazing along the shore meadows which attracted them? Another factor should also be pointed out: The coastline also hosted an archipelago, with a protective row of islands located offshore as seen today in the Netherlands and Northern Germany and these provided opportunities for closer contacts with the latter areas.

Jørgen Lund & Poul Nissen
Moesgård Museum






Lund, J., & Nissen, P. (2012). Alrum – Brandtomter i en vestjysk byhøj fra ældre jernalder. Kuml, 61(61), 75–109. Hentet fra