The passage of axes: fire transformation of flint objects in the Neolithic of southern Sweden.

LARSSON, LARS

Introduction

Studies of the circumstances surrounding the creation of material culture, as well as of technology and craft organization, are well represented in research. However, there is a need to study the causes of the destruction of material culture -- use, deposition in graves, caches, offerings, etc., as well. These make up an important part of the life cycle of an artefact, and this aspect is often perceived as self-evident. It is obvious enough in cases where a tool, after a period of use, loses its primary function. But the question is how often this really happens? There may have been many socially related reasons why an object was taken out of service, reasons associated not just with artefacts that were used, for example, as grave goods.

There are several aspects which may be considered within the processes of deposition and destruction. The large number of finds, mainly thousands of flint axes, in wetland environments in southern Scandinavia has been interpreted as an expression of ritual depositions (Bradley 1990: 57-64; Karsten 1994; Koch 1998: 158-61).

Based on the interpretation of these wetland depositions, the same explanation has sometimes been adduced about find circumstances connected to dry-land depositions with or without any connection to settlement sites. Deposition and destruction are combined when tools have been affected by intense fire. Intentional destruction of tools by fire is rarely observed among the bog offerings.

At settlement sites tools and debitage show traces of fire. However, our interest is aroused when one type of artefact shows a considerably higher percentage of fire damage than on other. This may indicate a conscious act of damage by fire. At some settlement sites there is an extraordinarily high percentage of fragments damaged by fire from finished, and often polished, axes in relation to fragments of other tools (Karsten 1994: 157-62; Apel et al. 1997: 39-41). Similar phenomena appear outside the entrances of megalithic tombs as well as in causewayed enclosures (Madsen 1997: 79; Andersen 1999: 41). The fragmentation of flint axes has been regarded as the result of intentional destruction in ritual acts in which fire played an important role (Andersen 1997: 300). The phenomenon of one or a few axes destroyed by fire is well exemplified, but massive destruction of large numbers of artefacts seems to be rare.

Sites with burnt axes

During the extensive revised survey of monuments carried out in Scania during the 1990s, fire-damaged flint artefacts were spotted in a field at Kverrestad in southeastern Scania, southernmost Sweden, about 15 km from the sea (FIGURE 1). A river valley runs from east to west, delimited to the south by a marked ridge (FIGURE 2). The site is situated on a small plateau on the southern side of the river valley (FIGURE 3) with marked slopes on three sides and with a wetland zone below the site. The bedrock at the site is made up of slate. In the main part of the excavated area the bedrock is covered only by a thin layer of clayey sand which equals the plough zone.

[Figures 1-3 ILLUSTRATION OMITTED]

Fire-damaged flints were found on the plateau within an area of approximately 70x70 m (FIGURE 4). The site was surveyed on several occasions, when every find was recorded in order to determine patterns in the spatial distribution.

[Figure 4 ILLUSTRATION OMITTED]

Excavations were carried out at Kverrestad in 1998 and 1999. They revealed a number of pits of varying size and depth in which damaged flint and stone artefacts had been deposited together with a considerable amount of fragmentary pottery (FIGURE 5). Burnt human bones, intentionally cracked into small pieces before deposition, were also found. The fragmentary state of the isolated bones as well as other finds throughout the filling show that they were deliberately fragmented before deposition during the entire process of filling in the pits.

[Figure 5 ILLUSTRATION OMITTED]

The distribution of surface finds does not coincide with the spread of pits filled with deposition remains (FIGURE 4). These pits had been cut into the soft bedrock and the contents protected during later ploughing. At the northern edge of the site an area was protected from deep ploughing, by large stones in the moraine just below the present surface. In this part a layer with a mixture of artefacts, probably part of the Neolithic surface, has been preserved. Most artefacts were found in small pits which seem to have been dug just 10-20 cm below the original surface. In other parts of the site similar shallow pits were destroyed by ploughing at least hundreds, perhaps thousands of years ago. Remains from these pits are now visible as finds in the plough zone.

Fragments from about a hundred thick-butted hollow-ground axes and chisels (FIGURE 6:1-2) have been found, as well as from a small number of thin-bladed axes. There is some variety in the degree of final polishing of the axes. Axes in earlier manufacturing stages without polishing are also found. Among the arrowheads, tanged arrowheads of a late type have been identified (FIGURE 6:4), as well as pressure-flaked projectile points. Some are leaf-shaped (FIGURE 6:5) while others have a marked tang (FIGURE 6:6). These types of points have not been identified in southern Sweden before. However, a small number have been found in Denmark (Ebbesen 1980) and explained as imports from the Oder area or further to the south. They are present in the Battle Axe Culture of the upper Oder area, north-eastern Germany. The finds from Kverrestad make up the largest collection of such points in Scandinavia. Fragments of a prototype daggers have been identified as well (Nielsen 1976) (FIGURE 6:10). Flake scrapers (FIGURE 6:11) and large blades (FIGURE 6:12) were also damaged by fire. Non-flint tools such as thick-butted axes (FIGURE 6:3) and battle axes -- regional types as well as examples from the west Danish Single Grave Culture (FIGURE 6:7) -- have been exposed to fire. Cracks due to intense heat facilitated later fragmentation of these stone tools. The stone axes have not only been fragmented, but were also subjected to acts of demolition, such as the destruction of the edge (FIGURE 6:3). Potsherds, mainly from vessels of a semicircular shape decorated with linear rows of twisted cord at the mouth of the vessel combined with large zigzag and herringbone design mainly formed by fine or coarse stamp impressions on the belly, are also present (FIGURE 6:8-9).

[Figure 6 ILLUSTRATION OMITTED]

A dating to the latest part of the Battle Axe Culture is in good agreement with all the finds. A radiocarbon determination from one of the pits, 3860 [+ or -] 75 BP (Ua-14883) c. 2300 cal BC, coincides well with other dates from sites with similar finds (Vandkilde 1996: 166).

Of special interest is the degree of exposure to fire shown by the different tool types. While more than 90% of the axe finds display changes by fire, two-thirds of the scrapers, half the number of the tanged arrowheads and one-third of the arrowheads made by pressure-flaking have the same alteration by fire. These marked differences indicate an intentional selection of which type of tool to burn. However, apart from a number of arrowheads, all the tools are in a fragmentary state.

No other site from the same period in southern Scandinavia has an artefacts component equal to Kverrestad. However, another site with massive fire destruction has been found in southeastern Scania. During surveys for Neolithic settlement, as part of research involving the cultural landscape in the 1980s, one such site was found at Svartskylle (FIGURE 1). Several thin-butted flint axes, a small number of thin-bladed axes, flake scrapers and blades had been fragmented by fire. The axe types date the finds to the later part of the Early Neolithic, c. 3400 cal BC (Larsson 1989). The burnt flints lay in four concentrations on a hill with a dominant position in the landscape.

Svartskylle is situated 17 km west of Kverrestad and 12 km from the coast (FIGURE 1). During the Early Neolithic this area as well as the coastal area bore evidence of well-established settlement (Larsson 1991).

Transformation by fire

Destruction of tools by fire is relatively frequent at sites from the Early to the Late Neolithic periods (Karsten 1994). The destruction of flint artefacts by fire seems to be less common in burial contexts dated to the Early and Middle Neolithic. However, during a late part of the Funnel Beaker Culture axes destroyed by fire are found outside the entrances of megalithic tombs.

There are few examples of destruction by fire in the context of burials from the preceding Battle Axe Culture (Lindstrom 1994: 61-4; Malmer 1962: 223-8; Sarlvik & Jonsater 1974: 93-4). Despite the presence of cremated human bones in Kverrestad, the deposition cannot be regarded as a kind of grave deposition. No other grave-good depositions show any similarities to the number of tools and vessels found. Another indication is the absence of unpolished axes in graves, since several axes deposited in Kverrestad were unpolished, just like axes in other kinds of offerings.

Ritually significant burning is evident in connection with mortuary buildings (Larsson 1988: 91) as well as at settlement sites of the Battle Axe Culture (Larsson 1992: 106). This practice continued until the latest part of the Neolithic. For example, two Late Neolithic flint daggers were found in a post-hole of a house at the Danish site of Oster Nibstrup, one of them with traces of fire (Michaelsen 1987).

Alteration of flint by fire is practised in other circumstances. Large quantities of unworked flint are often found in connection with megalithic tombs. Burnt flint is found mixed with clay which covers the orthostats and sometimes the entire chamber. In these cases the flint has been interpreted as constituting a material with a special quality, namely, the ability to absorb moisture in order to keep the chamber dry (Stromberg 1971: 228). Another purpose relates to an aesthetic quality of burnt flint -- the white colour. A thick layer of burnt flints covered the surface along the southern side of a passage grave, Kong Svends Hoj, on the Danish island of Lolland. Based on the amount of finds from some test pits, between 4 and 5 tons of burnt flints were used (Dehn et al. 1995: 95).

Although the time difference between Svartskylle and Kverrestad is more than a millennium, there are other finds in the neighbourhood from the Middle Neolithic Funnel Beaker culture which might partly bridge the time gap. Outside two megalithic tombs in southeastern Scania, features have been excavated which included flint tools exposed to fire. Eleven small concentrations of burnt human bones, including axes and chisels damaged or destroyed by fire, were found on the stone setting in front of the Trollasten dolmen (Stromberg 1968: 119-36). In some cases parts of the same axe were found in different concentrations. Some axes had been broken into pieces but were not altered by fire. Arrowheads and blades were exposed to fire. Nine small pits outside the passage of the megalithic tomb of Ramshog contained a filling of soot, cremated human bones and burnt flint artefacts (Stromberg 1971: 116-20). Some pits were found to contain fragments of axes destroyed by fire.

At both sites the majority of axes are thin-bladed. These as well as other finds date the features to the late Funnel Beaker Culture. The features and their contents are interpreted as remains of rituals. They show that the activities which are represented in Kverrestad were based on a tradition that belongs to the late Funnel Beaker Culture. Similar finds are lacking from the earlier use of megalithic tombs.

Burning experiment with flint

In order to find out what happens to flint in contact with fire, experiments were carried out. Axes of the thin-butted type equal to the axe type found at Svartskylle as well as thick-butted axes and chisels similar to the finds from Kverrestad were manufactured. The flint used was Senonian flint originating from southeastern Denmark, the same type from which the artefacts from Svartskylle and Kverrestad were made. These were exposed to different forms of fire, including direct as well as indirect heat in large and small fires, in order to find out what happened to flint at different temperatures. Before the experiments some axes had been stored in water for weeks in order to find out whether wet and dry axes reacted differently to fire exposure. When heated by fire, flint started to crack into pieces at a temperature of 270 [degrees] C. However, no changes of colour were recognized at that temperature. A clear change to white was documented at a temperature of more than 600 [degrees] C. Axes were placed at different levels in large pyres which were set on fire. In fires that exceeded a temperature of about 1000 [degrees] C the destruction of the axes was total. Fragments from such destruction have been found at Kverrestad. But almost intact axes or large pieces of axes with a white colour seem to indicate that the most important result was not to fragment the axes but to attain a change to a white colour. In order to achieve this transformation, axes were exposed to ordinary heat treatment (Olausson & Larsson 1982). Several hours later the axes were placed on a layer of burning charcoal with a temperature of about 600 [degrees] C. Gradually the colour changed. Small fissures occurred but the axes were not cracked into pieces (FIGURE 7). This shows that several axes, probably the majority, were handled carefully in order to obtain a colour change with a minimum of destruction.

[Figure 7 ILLUSTRATION OMITTED]

The social position of fire

What is being expressed at Svartskylle and Kverrestad differs from the destruction of single artefacts by fire evidenced at settlement sites, megalithic tombs and causewayed enclosures. This type of offering is found throughout the Neolithic, but on certain occasions the act achieves an imposing effect. Some wetlands depositions contain a large number of flint artefacts but these seem to originate from several small depositions over a period of several centuries (Karsten 1994: 129-133).

The cosmology which dictated burning, just like that related to wetland offerings, was current throughout most of the Neolithic. The fact that the offerings at Svartskylle and Kverrestad are the result of short-term activities indicates that they should be regarded as completed depositions of mass material. These seem to have been of exceptional size and intended to impress human as well as supernational beings. The destruction of material culture would have been very obvious and the wealth represented by the number of tools and exotics included must have been considerable.

The sites at Kverrestad and the earlier one at Svartskylle should not be seen as remains of ritual activities without any connection to other aspects of social life. The role of fire is evidenced from causewayed enclosures and megalithic tombs. So too is the habit of covering the remains of the acts of fragmentation, as seen in the depositions in ditches of the enclosures and by stone pavements at megalithic tombs (Stromberg 1968). The pits at Kverrestad are not arranged in a specific structure but they might be the last reminiscences of actions leading back to those in connection with causewayed enclosures.

Pits with a content of clay and with holes from posts were found on the highest ground at Kverrestad (FIGURE 5). Clay was not found in the immediate vicinity, so the filling had to be transported for a distance of at least a couple of hundred metres. They were arranged in a bow which might have delimited a certain area. However, no finds were made in the filling so the relation between these pits and the ones with Late Neolithic finds cannot be proved.

In addition, just a few post-holes were found within the excavated area. However, one has to be aware of the problems that the brown-coloured subsoil of decaying slate made it very difficult to discern post-holes and thereby traces of delimitations such as palisades. Something like a `second generation' of enclosures has recently been recorded from Scania and the island of Bornholm southeast of Scania (Svensson 1991; Nielsen 1999: 153) with similarities to henge monuments on the continent and Great Britain. They are dated to the latest Funnel Beaker culture and the Battle Axe culture. A site like Kverrestad, with something similar to a partial palisade, connected to pits with clay, might be one of several different types of sites of ritual importance connected to the later part of the Neolithic.

This type of public sacrifice of rare objects and with a more direct effect may have been practised on special occasions, probably in combination with external or internal threats. It could be an act which was primarily meant to legitimate power by impressing representatives of another community. The colour change of flint from the natural black or grey to white might be connected to a rite de passage. This change could be linked to the process when a human being is cremated, which results in fragmentary white bones. In that sense the use of fire on axes could be regarded as the cremation of the flint objects.

Conclusion

Fire played a much larger role in a variety of ritual acts during the Neolithic than was previously thought. In the selection of artefacts to be burned, with a predominance of axes, there is no important difference in relation to artefacts in wetland finds of ritual significance. Intentional destruction of flint artefacts by fire occurs throughout the Neolithic. In most instances we recognize this phenomenon through a higher frequency of fire damage in one type of artefact at a site or in a remarkably high percentage of axe fragments which are affected by fire. In a few instances, such as at Kverrestad and Svartskylle, a large number of tools had been destroyed at a special event. As no other similar sites are known, the question is whether these two sites are evidence of a special ritual behaviour in southeastern Scania or if similar sites have not yet been recognized elsewhere. Burning seems to have been involved in ritual acts which included external effects on the viewers as well as on the metaphysical world, a transformation of wealth with the hope of a direct and obvious effect for the community.

Acknowledgements. I would like to thank Mr Oltimo Andersson, the finder of the site, who has carried out the surveys of the site and took an active part in the excavation as well as the burning experiments. The landowner, Nils Jonsson, was of great help in the research. I would also like to thank Mr Arne Sjostrom, the leader of the excavation, as well as the excavation crew for their excellent performance. The excavations were funded by Hagglunds Stiftelse, Ebbe Kochs Stiftelse and Crafoordska Stiftelsen. The experiment with burning took part at the Lejre Research Centre, central Zealand, Denmark, which also funded that research. The bone finds were examined by Dr Torbjorn Ahlstrom.

References

ANDERSEN, N.H. 1997. The Sarup Enclosures. The Funnel Beaker Culture of the Sarup site including two causewayed camps compared to the contemporary settlements in the area and other European enclosures, Aarhus: Aarhus University Press. Jutland Archaeological Society Publications 33:1.

1999. Saruppladsen. Sarup vol. 2. Aarhus: Aarhus University Press. Jutland Archaeological Society Publications 33:2.

APEL, J., C. HADEVIK & L. SUNDSTROM. 1997. Burning down the house. The transformational use of fire and other aspects of an Early Neolithic TRB site in eastern central Sweden, Tor 29: 5-47.

BERGLUND, B.E. 1991a. The Early Neolithic landscape. Landscape, land-use, and vegetation, in Berglund (ed.): 6970.

(Ed.) 1991b. The cultural landscape during 6000 years in southern Sweden -- the Ystad Project. Copenhagen: Munksgaard International. Ecological Bulletins 41.

BRADLEY, R. 1990. The passage of arms. An archaeological analysis of prehistoric hoards and votive deposits. Cambridge: Cambridge University Press.

DEHN, T., S. HANSEN & F. KAUL. 1995. Kong Svends Hoj. Restaureringer og undersogelser pa Lolland 1991. Kobenhavn: Nationalmuseet. Stenaldergrave i Danmark 1.

EBBESEN, K. 1980. En Oldtidshoj ved Systofte, Lolland-Falsters Stiftsmuseums arsskrift 1980: 13-33.

KARSTEN, P. 1994. Att kasta yxan i sjon. En studie over rituell tradition och forandring utifran skanska neolitiska offerfynd. Lund: Almqvist & Wiksell International. Acta Archaeologica Lundensia 8:23.

KOCH, E. 1998. Neolithic bog pots from Zealand, Mon, Lolland and Falster. Kobenhavn: Der Konglige Nordiske Oldskriftselskab. Nordiske fortidsminder serie B:16.

LARSSON, L. 1988. Mortuary building above Stone Age grave. A grave from the Battle Axe Culture at Ullstorp, Southern Scania, Sweden, Papers of the Archaeological Institute University of Lund 1987-1988 n.s. 7: 81-98.

1989. Brandopfer. Der fruhneolithische Fundplatz Svartskylle im sudlichen Schonen, Schweden, Acta Archaeologica 59: 143-53.

1992. Settlement and environment during the Middle Neolithic and Late Neolithic, in L. Larsson, J. Callmer & B. Stjernquist (ed.), The Archaeology of the cultural landscape. Field work and research in a south Swedish rural region: 91-159. Lund: Almqvist & Wiksell International. Acta Archaeologica Lundensia 4:19.

LARSSON, M. 1991. The Early Neolithic landscape. Settlement and society, in Berglund (ed.): 70-72.

LINDSTROM, J. 1994. Glantan -- dodshuset fran stridsyxetid, Sormlandsbygden 1995: 59-70.

MADSEN, T. 1997. Ideology and social structure in the earlier Neolithic of south Scandinavia. A view from the sources, Analecta Praehistorica Leidensia 29: 75-81.

MALMER. M.P. 1962. Jungneolithische Studien. Lund: Almqvist & Wiksell International. Acta Archaeologica Lundensia 8:2.

MICHAELSEN, K.K. 1987. Oster Nibstrup, Danmarks laengste udgravning, Arkaaologi pti naturgassens vej 1979-86: 237. Kobenhavn: Poul Kristensens Forlag.

NIELSEN, P.O. 1976. De sfikaldte `madknive' af flint, Aarboger for nordisk Oldkyndighed og Historie 1974: 104-18.

1999. Limensgard and Grodbygard. Settlements with house remains from the Early, Middle and Late Neolithic on Bornholm, in C. Fabech & J. Ringtved (ed.), Settlement and landscape. Proceedings of a conference in Arhus, Denmark, May 4-7 1998: 149-65. Aarhus: Aarhus University Press.

OLAUSSON, D. & L. LARSSON. 1982. Testing for the presence of thermal pretreatment of flint in the Mesolithic and Neolithic of Sweden, Journal of Archaeological Science 9: 275-85.

STROMBERG, M. 1968. Der Dolmen Trollasten in St Kopinge, Schonen. Lund: Almqvist & Wiksell International. Acta Archaeologica Lundensia 8:7.

1971. Die Megalithgraber yon Hagestad. Zur Problematik yon Grabbauten und Gravbriten. Lund: Almqvist & Wiksell International. Acta Archaeologica Lundensia 8:9.

SVENSSON, M. 1991. A palisade enclosure in southwest Scania -- a site from the Battle Axe Culture, in K. Jennbert, L. Larsson, R. Petre & B. Wyszomirska-Werbart (eds), Regions and reflections. In honour of Marta Stromberg: 97-109. Lund: Almqvist & Wiksell International. Acta Archaeologica Lundensia 8:20.

SARLVIK, I. & M. JONSATER. 1974. Prastgardskulle i Veddinge. Undersokning av en forhistorisk gravhog, Varbergs Museums arsbok 1974: 85-94.

VANDKILDE, H. 1996. From stone to bronze. The metalwork of the Late Neolithic and earliest Bronze Age in Denmark. Aarhus: Aarhus University Press. Jutland Archaeological Society Publications 32.

LARS LARSSON, Institute of Archaeology, University of Lund, Sandgatan 1, S-223 50 Lund, Sweden. Lars.Larsson@ark.lu.se

Received 12 February 2000, accepted 14 March 2000, revised 15 May 2000.