Lystrup Østergård – En værkstedsplads fra yngre stenalder

Resumé

Lystrup Østergård
A Neolithic workshop

In 2007 a remarkable small site dating from the later part of the Early Neolithic Funnel Beaker culture was discovered at Lystrup, north of Aarhus in Eastern Jutland. Careful total excavation of the site revealed a well-defined cultural deposit with dense concentrations of flint debitage and implements lying in situ in a shallow hollow resulting from a group of windthrows. Via a series of analyses, the distribution of the finds relative to the individual features which were demonstrated, including a central hearth, has made it possible to reconstruct the events which took place, thereby permitting a detailed characterisation of the site. It can be perceived as a workshop which lay isolated in the landscape, at some distance from the actual settlement areas.

The finds primarily reflect two activities at the site: The production of blanks for thin-butted flint axes, where the raw material was obtained from the local moraine clay, and an activity which probably involved the working of bone or antler, judging from the remarkable number of burins which were recovered. Flake scrapers, normally the commonest tool type at settlements of this period, were virtually absent. The marked occurrence of burins and the site’s potential with respect to finds-distribution analysis together constitute a situation rarely encountered in a Neolithic context. Through identification of the various sequences of events, the activities acquire the character of brief targeted incidents.

Analyses of small Neolithic sites and the identification of specialised workshops can make a significant contribution to our understanding of the period’s patterns of ­settlement and activity – and prompt a critical examination of the settlement models used for the period to date.

In a series of regional investigations of Funnel Beaker culture settlement over the last three decades, a model has been applied whereby the sites are divided up into base settlements and hunting stations, respectively. Associated with these were sites related to the ritual sphere: offerings, graves and causewayed enclosures. Collectively, these reveal the general organisation of the population in the landscape. But this model is constructed on the basis of a general consideration that has, in particular, demonstrated changes in the settlement development through time. If we take a closer look at the individual sites in order to obtain a better understanding of the dynamics of  Neolithic settlement, the pattern of ‘base settlements and hunting stations’ becomes too rigid to work with – and perhaps even misleading.

Landscape
The site lies within the broad Egå valley, formed at the end of the Ice Age by glacial erosion and melt water. In Atlantic times a 5.5 km long and 1.5 km wide fjord, Egå Fjord, extended inland from the Bay of Aarhus. In the Early Sub-Boreal, when the site was active, the fjord was partially closed at its mouth by beach ridges and constituted a sheltered, shallow brackish water environment. The site lay on a low undulating moraine surface at the foot of the hilly northern side of the valley, 0.8 km from the shore of the fjord (fig. 1).

The site
On the edge of a slight elevation running down towards a narrow, peat-filled depression, the Neolithic finds extended over an area of 47 m² – 12 m in length and 5.5 m in width – within a shallow hollow (figs. 2-4). The hollow was characterised by two crescent-shaped features (A16, A41), together with a further oval feature to the north (A8) (see figs. 5-6). The crescent-shaped features were able to shed some light on the formation of the hollow. Their form and stratigraphy revealed that they resulted from windthrows. In an archaeological context this phenomenon is often connected with disturbances that have disrupted the stratigraphy of archaeological deposits. But in this case the trees had been blown over prior to formation of the archaeological deposits and the shallow root pits functioned as an actual surface for the activities.

The archaeological deposits had an average thickness of 5 cm and comprised dark, charcoal-rich sandy clay within which there was an even spread of dense finds concentrations. These lay directly over the heavy, stony yellow moraine clay. Intervening layers, for example earlier vegetation horizons, were not encountered. In certain areas, the finds, first and foremost flint debitage, a number of flint tools and a small quantity of potsherds, lay densely concentrated in up to three layers, one above the other. In some instances, heaps of homogeneous flint and axe flakes and chips could be readily distinguished, giving the impression of relatively undisturbed episodes of flint working.

In spite of a careful search, no traces of post-built dwellings were found associated with the archaeological deposits. A central hearth was, however, revealed as well as possible traces of a fence or a flimsy hut wall (figs. 9-10). The distribution and composition of the finds around the hearth revealed this to be the hub of the site’s structure, where various activities had taken place. Close to the hearth were two large stones which could have served as seats or work surfaces. An elongated flat-bottomed pit of uncertain function located directly north of the hearth should perhaps, together with the discovery of a polygonal axe, be seen as an indication of ritual activities.

The orientation of the windthrow pits shows that the trees fell away from each other, resulting in the formation at the site of a small sheltered hollow with exposed moraine clay (figs.7-8). The site stratigraphy suggests that only a short period of time elapsed before the finds were deposited within this hollow. This observation prompts the article’s hypothesis that the windthrow pits gave access to the moraine clay’s rich content of flint, which was then worked in situ.

The artefacts
The artefacts are predominantly of flint. In their manufacture, use was made of local moraine-deposited flint which in this area is of particularly good quality and varied type. The flint tools and flint debitage have a total weight of 74.1 kg; the tools number 295 examples, while the debitage is estimated to include c. 10,000 pieces.

The distribution of flint and stone artefacts is given in tables 1-2. Almost half the flint debitage can be linked to the production of axes of thin-butted type. In addition to large quantities of various axe flakes/chips, there are seven discarded blanks/rough-outs and 20 hammerstones (fig. 16).

In the tool inventory, special attention should be drawn to the 120 burins (40.7%), an unusual feature in a Neolithic context (figs. 11-14), in addition to 38 core and flake drills (12.9%) and 35 knives (11.9%) (fig. 15). Further to these are 52 small tools in the form of blades or flakes with retouch or visible use-wear (17.6%). The burins were produced on simple robust flakes that appear to have been specially produced for the purpose. Transverse burins on retouch are in the majority, followed by edge burins (table 3). One find stands out from the rest, namely half of a finely-worked polygonal axe of basalt (fig. 17). This was not made at the site.

A small, poorly-preserved assemblage of pottery (1.8 kg) lay deposited in concentrations around the site. In terms of vessel forms, the presence has been demonstrated of funnel beakers, a lugged beaker and a bowl. The decoration is characterised by simple rim ornamentation, vertical belly stripes and the use of twisted cord (fig. 18). The minimum number of vessels represented in the assemblage is calculated to be seven.

Dating
The typological date for the site is based on the pottery, the flint axes, the polygonal axe, denticulates, a single ‘disk knife’ and, to a certain extent, the burins. The vessel form and decoration of the pottery corresponds to the Funnel Beaker culture’s phase TN II. There are close parallels in the pottery recovered from the palisade ditch at Sarup I, which is linked to the Fuchsberg group (note 38). This date is also supported by the flint and stone tools, although these also open up the possibility of a component from the subsequent MNA I.

A radiocarbon analysis of a charred seed coat from the archaeological deposits near the hearth shows, with a probability of 95.4% (±2 standard deviations), a double peak with an 8.6% probability of a date of between 3630 and 3580 BC and a 86.8% probability for 3530-3360 BC. The greatest part of the curve corresponds, accordingly, with the radiocarbon dates for Sarup I.

With a possible small component from MNA I, the date for the archaeological deposits falls within the Funnel Beaker culture’s TN II phase with links to the Fuchsberg group.

Subsistence
The conditions for preservation of bone at the site were unfortunately very poor. The humus content of the archaeological deposits does, however, bear witness to the presence of a certain amount of degraded organic material. The animal remains comprise two badly-preserved teeth of, respectively, a young domestic cow and a large ruminant. Further to these, 11 small bones were found by fine sieving, of which three are fish bones, probably cod.

Soil samples processed by flotation yielded 23 charred cereal grains, of which 11 were of barley and one of wheat, while the others were unidentifiable. Charred hazelnut shells featured in several samples and a single charred apple pip was recorded.

A strange component of small water-rolled stones found in the deposits could possibly originate from seaweed, bladder wrack, gathered on the coast. The function of the seaweed is unclear, but there are a number of possibilities, e.g. a soft underlay, fuel, animal fodder or manure; it could also have constituted human food.

Activities and activity areas
The natural sources of good raw flint in Eastern Jutland are the coastal cliffs and potentially also the banks of streams and rivers, where the flint is exposed naturally and can be gathered directly. On the forest floor of the interior, flint would have been rarely encountered. It seems therefore very likely that the hollow created by the windfalls gave very welcome access to the flint in the moraine deposits, which could then have been the subject of more systematic searches and collection. Several of the flint nodules found in the archaeological deposits have only one or a few scars resulting from blows, probably resulting from testing of the flint quality. One very large block (42 kg) was found in four pieces scattered around the site, with a few missing pieces that could have been worked further (fig. 19).

The debitage from the axe production has been analysed with the aim of discovering the types and number of axes produced at the site. Several definite axe-knapping episodes have been distinguished on the basis of in situ concentrations, identification of debitage from the same flint nodules and with the aid of refitting (figs. 21-23, table 4). The flint flakes have been classified according to the use of hard and soft knapping techniques, i.e. the employment of, respectively, hammerstones and fabricators of antler, in order to discover the number of stages in the production of the four-sided axes present at the site (figs. 20, 24-27). In the course of this analysis the character and extent of the material was compared with related finds and the results of modern experiments (note 60).

Large flakes retaining the original cortex of the flint show that some pieces were produced in situ from raw unworked flint nodules (stage I), whereas other examples appear to have been brought to the site as roughly-worked axe blanks (stage II). The aim of the production was the manufacture of axes up to stage III. No clear traces of stage IV, the last trimming of the axe sides and edges, or of the final polishing, stage V, could be demonstrated. A total of about 15 individual axes were worked at the site, of which about half were abandoned and discarded at the site as failures, while the finished examples were taken away to another workshop or a base settlement to be given their final finish. Through comparisons with modern experiments, the total time expenditure for the axe production is estimated as a maximum of 12 hours. If production was continuous, then all that was involved was a single day’s work for two flint knappers. The quality of the work is considered to be fully on a par with the general level in the Funnel Beaker culture.

The other activity that characterises the site is apparent from the large number of burins in the assemblage. Burins are associated with the working of hard materials such as antler and bone, and this was confirmed by wear analysis of 13 pieces from the site. The activity could well have involved other elements of the inventory such as drills, knives and diverse tools with retouch. The activities took place in particular in the vicinity of the hearth, but a particularly high concentration of burins and burin spalls was found on the eastern periphery, in the deeper part of the hollow, behind a possible fence (fig. 28). This could represent the deposition of burin waste or the existence of a small isolated work place.

Even though burins rarely occur in large numbers at the settlements of the period, they are occasionally present and in a few cases they are seen in large numbers as for example at the site of Grønvang 2, near Kalundborg in Western Zealand.

The items which were produced could have been antler axes, chisels, bodkins or harpoons. A close relationship with the production of flint axes is also conceivable in the form of the manufacture of antler fabricators. This is, however, not supported by evidence from other flint axe workshops, where burins have never been recorded in the tool inventory.

The settlement around Egå Fjord in TN II (- MNA I)
The area around the site and along the northern side of the fjord has, over the course of the past 12 years, been subjected to extensive and comprehensive archaeological investigation in connection with road construction and development of building land. It is therefore now possible to see the site in a wider settlement-related perspective for the period TN II - MNA I (fig. 29). The nearest settlement-like finds have been located 325 m ENE of the workshop site, but these are difficult to evaluate in detail due to disturbance later in prehistory. Possible base settlements with the remains of houses were encountered 2.1 km north and 7 km west of the site, respectively. In addition, possible hunting stations were demonstrated on the nearby shore of the fjord. Four other sites within a 2 km radius bear witness to ritual activities; these comprise two isolated system-ditch complexes and two dolmen sites.

The area within a radius of 300 m of the site has been investigated via field-walking and trial excavations, and these did not reveal the existence of any contemporaneous settlement traces here. It can therefore be reasonably securely concluded that the workshop lay at a distance from the settlement sites. It is possible that it was located on the edge of recently-established arable fields. Clearance of the primeval forest would have given the wind easy access to the old forest trees which then, at the woodland edge, became easy victims for storms.

Workshop sites of the Funnel Beaker culture
During the Funnel Beaker culture, workshops were often associated with flint quarrying and flint-knapping sites and several of these were specifically oriented towards axe production, for example that at Hastrup Vænget in Eastern Zealand.

Apart from axe production, specialised workshop activities have rarely been recognised in the Funnel Beaker culture. The above-mentioned Grønvang 2 on Zealand resembles Lystrup Østergård with respect to its size and a large content of burins. Another site with a specialised activity is Studeli Klit in Northern Jutland, characterised by a huge number of flake drills.

Neolithic sites that were not actual ordinary settlements but sites for special workshop activities are possibly under-represented in the overall archaeological record, either because they are small and easily overlooked during archaeological investigations or because their uniform and more specific site circumstances are more vulnerable to repeated and possibly also changing use of the localities. Several of the sites we perceive as base settlements could possibly represent the accumulated remains of more specialised activities. An important feature type relative to so-called base settlements is the house! Investigations of Scandinavian house remains from the period have demonstrated a clear tendency for houses, activity areas and refuse deposits not to be located in the same place; there may possibly have been rules with respect to cleanliness around settlement areas. This tendency has subsequently been demonstrated in connection with new archaeological investigations in Scania and in the Sarup area in SW Funen.

Consequently, we must see settlement and activities in the early agricultural society as a more widespread and dynamic use of the landscape. In future regional investigations it will be important to look critically at the term ‘settlement’ and distinguish to a greater degree between sites for activities, refuse deposition and habitation. During excavations we should be aware of the minor find complexes and focus on their possible unique features – and remember that houses are to be looked for at some distance from the find-rich areas.

Uffe Rasmussen
Moesgård Museum