Vuollerim Six Thousand and Fifteen Years Ago

The theoretical background that led to the discovery and excavation of the Late Mesolithic semi-subterranean dwelling at Vuollerim is presented. This is followed by an interpretation of the social function of the house. A comparison with contemporary dwellings is made and the cultural context of the Vuollerim site is explored. The paper ends with some observations and thoughts concerning the impact that the archaeological discovery in Vuollerim had on its present day inhabitants and how it changed their view of themselves, their history and their future.


INTRODUCTION
In my opinion, one ofthe most exciting results of the Lule River Interdisciplinary Research Project was the discovery ofa Late Mesolithic dwelling at Vuollerim in the Parish of Jokkmokk in northern Sweden. The site was discovered 15 years ago and the excavation of the house (commonly known as Norpan 2) was completed 11 years ago. Only a few publications have been written since then and there has been no overall presentation of the excavation in English since 1985, or in Swedish since 1986 (LoeAler & Westfal 1985Westfal , 1986. It need hardly be mentioned that both of these papers were obsolete within six months of publication due to the ongoing excavations. The absence of any recently published material concerning this site has led to the somewhat embarrassing situation in which antiquated papers are still being cited by colleagues. It is my intention to partly rectify this deficiency by briefly presenting the essentials concerning the excavations carried out between 1983-1987,accompanied by a few observations on the similarities and dissimilarities with contemporary Mesolithic sites in northern Sweden (fig. I). This is followed by a preliminary hypothesis concerning the social organization and settlement patterns that created the remains at Vuollerim and its cultural context. In conclusion, a commentary concerning the effects that the discovery has had in archaeology and on the historical awareness of the local community is presented. I am not an empiricist. I am quite aware ofthe connections and dependencies between what we consider as facts and prevailing theories. I am convinced that our theories determine both what is chosen and not chosen as relevant, either explicitly or implicitly, in any given scientific situation. Despite this awareness I am not going to elaborate on the possible reasons behind what is presented here, nor the how's and why's behind the interpretation of the material. Let it be sufficient to say that I fully expect this paper to become redundant within the near future thanks to the continuing work on Vuollerim. The ecological component of the model presumed that the hunter-gatherer groups would try to maximize their efforts by exploiting those environments which contained many diversified ecological niches within a relatively limited and well defined area. Environments containing rivers, streams, bogs, forests and lakes within a radius of 10 kilometers were considered as likely areas in which prehistoric societies would have situated their base camps.
The resulting archaeological survey was goal oriented. Its aim was to find and document prehistoric sites as specified by the theoretical models. Map studies were carried out with the theoretical concepts in mind and areas that seemed likely to fulfill the hypothetical requirements were targeted for archaeological survey.
The theoretical models said, among other things, that hunter-gathering societies should have moved within well defined territories and that they would have situated key sites in ecologically rich areas. Ifthis were so, then one would expect, sooner or later, to find the remains of substantial structures that served as winter dwellings. Exactly what such a structure would look like was any one's guess. Ethnographic studies showed that the most common winter dwelling in circumpolar areas was the subterranean or semi-subterranean dwelling. The work that was then being carried out on embankments of firecracked stone in central Norrland by Åsa Lundberg and Lars-Göran Spång was well known. They proposed that the embankments of fire-cracked stone were the remains of semi-subterranean winter dwellings and they were undertaking excavations to gain support for this hypothesis. Embankments of firecracked stone had not been found in northern Norrland, but the general trend of Lundberg's and Spång's reasoning was in accord with that of Forsberg's and Westfal's. Thus the main focus of the archaeological survey in the Jokkmokk area was to discover the remains of semi-subterranean structures, Current Swedish Archoeotogu, Vol. 7, l 999 Fig. 2. Fouz hvpothetical models that illustrate the fbur presumed settlement pattez ns that could have existed during different prehistoric periods in not thezn Norzlaml. The archaeological suz veys izz and around Jokl zzzokl pzzrish in l983 were carried ozzt wi tlz the intention offznding and doczzmenting sites tlzat zvozzld confirnz tlze models credibility: Tlze simplified illustrations above u&ez'e zedzawn fi ont photographs of' tlze oziginal drawizzgs as tlzez were pzesezzted in the l983 ezhibitiozz.
A) 7000-4000 BC. Small sites inhabited by small groups of' hunter-gatherers with a nzobile lifestz&le. B) 4000-2000 BC. Larger sites in the forest areas with u&i nter dwellings (base camps) occzzpied by& a local group witlz seasonal nzovements to temporary sites and activity areas by segments of the local grozzp. C) 2000-500 BC. Larger sites in the forest areas with winter dwellings (base canzps) with seasonal movements to temporary sites. Specialised hunting activi ties were carried out in the mountai n areas on a seasonal basis. D) 500 BC-500 AD. Sites grouped in distinct concentrations in the forest and mountain azeas with empty areas betzveen the different territories.  Fig. 6. /sometric map showing the distribution of the 2/ 373 grams of bone that wetefound on or in one of the three floor areas, in features or on the original ground surface. The distribution pattern of this material is bimodal, with two main concentrations on each side of the house. A third concentration is centred on and around the rubbish pit just north of the house. A total of 55, 444 grams of burned and unburned bone vvere recovered. The large amount ofbone recoveredis unusualfot sites in this part of the country. The preservation of the bone material is probably due to the fact that the lime content of the soil in this area is son&ewhat higher than normal fot this part of' the countpi. 51 m'. Studies concerning the relationship between dwelling and household sizes of northern hunter-gatherers suggests that houses with a floor area exceeding 20 m' were inhabited by more than one family (Helskog 1994;Lundberg 1997:100).This conclusion coincides with the bimodal pattern exhibited by the different material remains and, to a certain extent, the placement of the different features found within the boundaries of the main floor area.
The similarity between the distribution patterns of stone artefacts, quartz flakes, bone material, fire-cracked stones and red ocher is  The house can also be interpreted as a base camp. A base camp can be defined using either social or economic criteria (Forsberg 1985(Forsberg , 1988). There are two types of socially defined base camps. On the one hand there are small base camps which are inhabited by a local group consisting of4-8 families which is the ". ..fundamental socio-economic unit of hunter-gatherers inhabiting the northern taiga areas. .. " (Forsberg 1988:59).The other type is the large base camp of the microband, which is made up of a number oflocal groups with a population of approximately 100 people.
There are also two types of economically defined base camps: semi-permanent and permanent. The semi-permanent base camps are occupied seasonally. Societies thus organized have at least two or more semipermanent base camps which are located in resource rich environments. People move between these different camps in order to utilize the different resources as they become available during different times of the year.
Permanent base camps are also located in resource rich environments but these sites are more or less inhabited throughout the year.
Seasonally dictated movements from permanent base camps are only undertaken by parts of the population during the year. The size, duration and length of these movements is determined by the type of resource which is to be exploited.
Base camps of all types differ from other sites in that the archaeological remains on the former are more diverse and varied due to the fact that base camps are occupied over longer periods of time by larger numbers of people who are engaged in many different types of activities (Forsberg 1985(Forsberg , 1988.
The house in Vuollerim displayed a wide range of archaeological materials and remains. It is possible that the other three houses in the area were occupied at the same time and therefore together constitute the remains of a base camp for a local group. The features that have been interpreted as storage pits suggest that the house functioned as a permanent base camp in which seasonal resources were transported into the camp, rather than moving the total population to the seasonally available resources.

FEATURE A4a AND A23/27
During the excavation of the large semisubterranean house two other highly interesting features were uncovered which so far have never received any attention. Features A4a and A23/27 are two separate floor areas; the former was 4.5-5 meters long and 3.5-4 meters wide and rested directly on the original ground surface. The latter was 3.8 x 3 meters in diameter and 10-20 centimeters below the original ground surface. One cooking pit and one posthole were discovered within the boundaries of feature A4a and three postholes were uncovered in feature A23/27. No cooking pit was found within feature A23/27 but one was discovered about 2 meters north of it. These two features have been interpreted as the remains of two tent like constructions.
Both contained tools, bones, fire-cracked stones, flakes and red ocher.
Cnrrent Swedish Archoeology, Vol. 7, /999 Feature A23/27 diAers radically from the other floor areas in that it contained abundant amounts of red ocher powder as did one of its postholes. The posthole in feature A4a was similar in that it also contained a 2 centimeter thick bottom layer of red ocher. Otherwise, feature A4a did not contain noticeable amounts of red ocher.

Hunter-gatherers of Siberia and North
America usually had both winter and summer houses. The thick earth and wooden walls and/or roof of the winter houses became uncomfortably damp during the summer which compelled its inhabitants to move into lighter and airier constructions (Lundberg 1997:95 plus references). This would be a plausible interpretation of these two tent like structures. But this does not explain the presence of the relatively high concentration of red ocher found within feature A23/27. This explanation assumes that the large house was contemporary with the tent constructions. Feature A4a was covered with soil from the embankment which surrounded the larger house, which implies that feature A4a is somewhat older. A radiocarbon date from the cooking pit in feature A4a gives some support to this assumption. The cooking pit was dated to 5575 BP which means that it could belong to the beginning of the site's occupation. Feature A23/27 was partly covered by soil from the embankment of the larger house and partly by soil material from its own embankment, which would indicate that it was in use during the later phase of the site.

EMBANKMENTS OF FIRE-CRACKED STONE AND VUOLLERIM
In order to gain insight into the possible cultural context of the Vuollerim dwelling it is necessary to compare it with other contemporary houses. Åsa Lundberg has in a recent dissertation shown that the feature commonly known as embankments of fire-cracked stone (Sw. skä rvstensvallar) are in fact the remains of semi-subterranean dwellings that were in use from the Late Mesolithic up until the end of the Neolithic (Lundberg 1997). A comparison between these inland dwellings and Vuollerim followed by a further comparison with the contemporary coastal dwellings at Lillberget in Överkalix parish, Norrbotten shows surprising similarities and differences.
The location of the Vuollerim house and its relationship to other features is similar to that of most embankments of fire-cracked stone. Embankments of fire-cracked stone are often located near the confluence oftwo water courses and are often placed 5-20 meters from the shore line (Lundberg 1997:102,123).The four houses in Vuollerim at the time of their occupation were probably located in a delta like landscape near the confluence ofthe Lilla and Stora Lule Rivers between 5-20 meters from the water's edge.
The house at Vuollerim, like most embankments of fire-cracked stone, is located on gentle sloping sandy ground with good drainage properties.
Two to five embankments of fire-cracked stone are usually found together, with an internal distance of 100-500 meters between them. The distance between the four houses in Vuollerim is 60-180 meters.
Lundberg suggests that the presence of moose was taken into consideration when choosing a dwelling site. Most embankments of fire-cracked stone are found in conjunction with hunting pits (Lundberg 1997:123). There are a number of hunting pits in the vicinity ofthe Vuollerim site and one has been radiocarbon dated to 6315+315 BP.
There are also many similarities concer- Vuollerim's floor area is rectangular in shape and about 51 m'.
The greatest diflerence between Vuollerim and its contemporaries is the size of the embankment which surrounds the floor area. The latter is often both thicker and wider than that which surrounds the house at Vuollerim.
As the name implies, embankments of firecracked stone are made up almost entirely of fire-cracked stones. This is in stark contrast to the embankment at Vuollerim which is mostly composed of sand. From one embankment in Vilhelmina the amount of firecracked stone for every excavated level of 10 centimeters per square meter was 50-100 liters (Lundberg 1997:50). In contrast, the embankment at Vuollerim produced on average 1-2.5 liters for every 10 centimeter per square meter. In both cases the shape and size of the cooking pits and hearths found inside and outside of the houses are almost identical: bowl shaped, between 40-70 centimeters in diameter and 30 centimeters deep (Lundberg 1997:108).And in both cases these features have not been in use simultaneously but were utilized, abandoned and cleared out at diflerent times.
Features that have been interpreted as storage pits have been found on sites with embankments of flre-cracked stone and at Vuollerim. But while the three storage pits at Vuollerim were located within the house itself those found on sites with embankments of fire-cracked stone were placed outside of the dwellings (Lundberg 1997:110).
In both cases quartz is by far the most common raw material. It is accompanied by varying, though often minor amounts of quartzite, flint, slate, greenstone, hälleflinta and sandstone. Scrapers are the dominate tool type on sites with embankments of fire- The amount of burnt and unburnt bone recovered at Vuollerim, about 55 kilograms, surpasses that which is usually found elsewhere in Norrland. A proper osteological analysis has not yet been undertaken. The limited work that has been done shows that the material contains bones ofmoose, beaver, fish and birds, an assemblage which is typical of that which has been recovered from embankments of flre-cracked stone.
The flake material from Vuollerim also surpasses that which is usually recovered from embankments of fire-cracked stone. From the former 36,952 flakes were recorded. This is in stark contrast to the 2000-8000 flakes which are usually recovered from embankments of fire-cracked stone, with the exception of Bellsås in Tåsjö parish, Jämtland where 25, 197 flakes were recorded (Lundberg 1997:26ff). Lundberg points out that this difference is probably the result of how accessible quartz was in relation to the sites.
In the case of Vuollerim there is a quartz quarry 10 kilometers upstream from the site which was presumably in use at this time (Falk 1995).
Lundberg calculates that embankments of fire-cracked stone were inhabited between 500-1600 years and that a house construction itself was rebuilt every 30-70 years (Lundberg analysis 12). It is possible that the river changed its course during the time of the site's occupation, causing the site to be abandoned, and that the high quality of preservation and the apparent short duration of the site is a consequence ofthis event.
LILLBERGET AND VUOLLERIM The Lillberget site, Raä 451, in Överkalix parish lies 11 kilometers northwest of Överkalix and 120 km east ofVuollerim. Lillberget is one of a number of similar sites that have recently been discovered in the coastal region (Baudou 1992:69).So far it is the only site of Current Swedish Archaeotogy, Vol. 7, 1999 its kind that has been documented and presented through the vigorous efforts of Halén (1994).A comparison between Lillberget and Vuollerim shows some similarities but the differences are even more striking. The Lillberget site lies on a gentle north slope composed of sand just above a bog. Today the site is situated between 58-64 m. a.s.l. but at the time of occupation, 3900 BC, it was located near the beach on a island at the bottom of a fjord like formation about 35 kilometers from the open sea.
The dwelling site consists of ten semisubterranean houses placed wall to wall in two groups. The floor area lies 40 cm beneath the surface of the ground and the houses are between 3.5-14 meters long and 3.5-9.6 meters wide (Halén 1994:83), a few being comparable in size to the house in Vuollerim.
Halén excavated three houses. Two hearths were found in each house, one on each side of the floor area. The hearths were rather large, being between 116-157 centimeters long and 57-124 centimeters wide (Halén 1994:86ffl. Quartz made up 79.5% of the stone material, followed by flint with 11.7% and greenstone, 7.6%. The remaining 1.2% is made up of sandstone, slate, hälleflinta, quartzite and jasper. Ninety-four tools of quartz were recovered: 65 scrapers, 12 blades, 10 cores, 5 drills, 1 knife and 1 transverse arrow. Of the 69 tools made from flint there were 33 scrapers, 14 bifacial points, 13 blades and 2 transverse arrows. Nineteen tools of greenstone were recovered, 7 of these were axes or chisels and 4 were blades. Four scrapers and a number of flakes of red and green banded jasper were also found. Only one artefact ofred slate was recovered, a small arrowhead with a wide tang and short straight barbs (Halén 1994:99A).
Sixteen kilograms of Comb Ware ceramics and a bead of copper were also recovered from the Lillberget site (Halén 1994:128, 153).
Fire-cracked stones were found on the site but in such small amounts that their presence has only been noted in passing.
An analysis of the 2106 grams of bone that was recovered shows that 81.1 % belong to different species of seal, 11. 7% to moose or reindeer and 4. 1% to beaver (Halén 1994:164).
Halén maintains that at least nine of the ten houses were inhabited at one and the same time with a population between 80-100people divided into 20 families. The site functioned as a permanent base camp which was in use throughout the year for a period of between 17-33 years (Halén 1994:175ff).
There is a certain amount of similarity between Vuollerim and Lillberget. Both sites share certain topographical features. The size of the houses is in some cases identical.
Quartz is the dominate stone material and scrapers are the most common tool type.
Thereafter, the differences loom larger than the similarities. The placing of the houses wall to wall, the relatively few hearths in each house, their large size, the lack of other types of features, the apparent and near total lack of fire-cracked stones, the relatively large amounts of tools made of flint and greenstone and the almost total lack of tools made from slate and quartzite, the dominance of seal bone with traces of reindeer, the ceramics, the copper bead, the transverse arrows and bifacial points and the short occupation period are all substantial differences between these two sites.
The presence of one flake of red jasper as well as the few flakes and five microblades of flint and possibly even the greenstone artefacts from Vuollerim could be the result of some form of contact with the Comb Ware complex along the coast of which Lillberget is a part. The single point of red slate with a wide tang and short straight barbs, of which a similar type was found at Vuollerim made of greenstone, could also be the outcome of some type of contact.
VUOLLERIM' S CULTURAL CONTEXT In contrast to Lillberget, Vuollerim appears to have many more similarities with the embankments offire-cracked stone. The same types and frequencies of the stone materials and tool types, the presumed composition of the bone remains, the topographical location of the sites, their relationship to other houses and other types of remains, the size of the houses, the types of features found in and around the houses and the size of these features are all particularly similar.
The overall impression and the pre- ofpeople. This shared cultural tradition can be defined as a social territory with ". ..one or more related language groups which share a common material culture. .. " with social relations and contacts in the form of movement of both people through marriage and products (Lundberg 1997:131).
The similarities between Vuollerim and the embankments of fire-cracked stone becomes even more apparent when compared with the coastal site of Lillberget with its relatively large portion of typically Finnish and Russian components. The limited similarities between the quartz materials found on these two sites shows that they probably belonged to a common techno-territory in which the social relationships are practically absent (see Lundberg 1997:131).
Downstream Vuollerim and along the coast there are many newly discovered but unexplored sites. Some of these, like Lillberget, may belong to the Comb Ware Ceramic tradition, while others might have been utilized by the Vuollerim population as seasonal camps and/or by related groups as seasonal and/or base camps. It is therefore difficult to judge the degree and type of social contact that took place between these two different cultural groups solely from the limited comparison between Lillberget and Vuollerim; a comparison which indicates that contact was restricted. This presumed lack, or minimal amount, of contact might reflect that they belonged to different cultural traditions and/or social territories.

THE LEGACY OF VUOLLERIM
The excavations at Vuollerim resulted in more than just the discovery of a semi-subterranean dwelling from the Late Mesolithic period.
Lars Forsberg's work resulted in a doctoral dissertation which elaborated on the four theoretical models concerning the settlement patterns in Upper Norrland between 1500 BC to 0 AD (Forsberg 1985). This was the first archaeological dissertation which utilized archaeological materials from this area. Never again could one look upon Upper Norrland as a vast empty wasteland periodically traversed by mobile hunting groups on their way to some other place. By applying concepts previously developed by Broadbent (1979)for the northern coastal regions, Forsberg could present a description ofdifferent, complex and well organized prehistoric societies which had developed well thought out and elaborate subsistence strategies which were highly integrated with the prevailing ecological conditions in this part of northern inland Sweden.
The discovery of the semi-subterranean dwelling at Vuollerim quickly led to similar discoveries. The nationwide archaeological survey conducted by The Central Board of National Antiquities (Sw. Riksantikvarieämbetet) reached Upper Norrland in 1984 (Klang 1986:8).The Vuollerim site was used as a training ground where the first survey workers could familiarize themselves with this new type of feature. It was not long before similar types of remains were discovered and registered. Today there are many hundreds of similar sites known throughout the area, both singly or in larger concentrations which may well be the remains of whole villages (Klang 1987 Further eastwards one sees the houses built for the workers and even one or two of the mess halls and bachelors' barracks.
At the end of the 1970s most of the dams and power plants were nearing completion. Unemployment grew and in its wake many personal tragedies started to unfold. Bank loans and mortgages defaulted. People lost their homes, shops and businesses. Alcoholism spread; families broke apart and people started to move away. The history of these people seemed to both start and end with the beginning and completion of the exploitation of the water resources. For a community which in all appearances did not posses a history longer nor more enduring than that which was associated with the technological intervention it was difficult to build a future which seemingly had already ende d.
The archaeological excavations at Vuollerim would have repercussions far outside the limits of the academic discipline itself.
Suddenly this rural community was in possession of a history which not only reached at least 6000 years back into the past, but into Norrland's past. The archaeological discovery made people aware of a different perspective concerning their own conceptions about their own time and their own future. Other people and other societies had existed long before their own and had done so independently of the exploitation of the hydroelectric power resources. These other and different communities had developed, flourished and disappeared to be replaced by something else. This deep awareness with its long term historical perspective woke many to the possibility that Vuollerim's and Norrland's Current Swedish Archaeolo~, VoL 7, 1999 destiny was not solely dependent on anonymous powers from without. If this part of Sweden possessed a 6000 year old history then it might well have an equally long future. The decline which the present community was suffering from was not a permanent state, nor was it a natural condition. It was a short period in a long chain of transformations.
This historical insight provided people with a new self awareness; that the potential for something new was already to be found within the region itself, in its own people, in themselves. Suddenly the community had both a past and a future, and the contents of that future were neither permanent nor predetermined. On the contrary, it could be modified and it could be transformed.
One of the most interesting results from the excavations at Vuollerim has never received any attention from archaeologists.
It was not the discovery of the remains of a unique dwelling (it's hardly that). Neither was it because it was one of the most ancient sites found in the area (which it is far from being) and not because it was the first to be found, The really interesting result from the excavation was that it had repercussions far beyond archaeological circles; that it served as a catalyst and introduced a long term historical awareness and perspective on contemporary events and present day society.
This happened independently of any apparent political agenda and for the most part beyond the boundaries of academic archaeology. A rural community awoke to the realization that history and prehistory are meaningful and that it could enrich their lives and understanding of both the past, present and future. This, I think, is the real legacy of Vuollerim.
Over 200 articles and 9 doctoral theses have been published in connection with the Lule River Research Project, some of which deal with different aspects ofthe river valley's recent history (see Baudou 1996 for references). But the fact remains that the archaeological work had a profound impact in ways which the other subjects involved in the project did not (Nordström 1989(Nordström , 1991.
Why? And how did the discovery in Vuollerim affect the Saami, the oldest known indigenous population of the area? Did it act as an integrating or segregating factor? Did it, or was it used, to further marginalise them and other minorities from mainstream Swedish society, either intentionally or unintentionally? Were there political intentions here that were not apparent in the beginning or of which I was not consciously aware of then?
These are sharp questions that need answering, but that would require a whole new article.