Culture Moves Like an Octopus Aspects on Archaeological Regions and Boundaries

The graves in the interior of central Norrland in northern Sweden exhibit a wide range of forms throughout the Iron Age. Burial mounds and stonesettings of various forms also occur in different environmental contexts. The purpose of this paper is to discuss the spatial and cultural significance of the Iron Age graves in the provinces of Jämtland and Härjedalen. The point at issue is what the distribution of so-called lake graves and burial mounds represent with respect to the general discussion of archaeological regions and cultural traditions. It is here suggested that the concept of region, when it is applied to material culture, represents one of the many aspects of cultural configurations, but as we are dealing here with a longterm perspective the cultural world is continuously in motion, which makes regional boundaries hard to grasp.


INTRODUCTION
Archaeological regions have long been central to the discussion of how we interpret the spatial distribution of the archaeological material, but what is a region and what is characteristic of its existence?The regional concept has during the 20'" century become the very essence of geographical philosophy, but it is important to note that the concept has been approached differently from time to time and that there are various kinds of regions to take into account (Helmfrid 1969).The classical use of the concept, the natural region or the a priori region, was in the early part of the 20'"century more or less defined by contrast and homogeneity.A new concept, cal led the functional region, was later introduced.This type of region put more emphasis on the functional connec- tions between objects and their relations to central areas, instead of focusing on various kinds of individual characteristics that appeared to show contrast between different geographical environments (Helmfrid 1969:18).
With regard to the discussion ofcreating a geographical division into "archaeological single graves that have been recorded by the Ancient Monument Survey.The proposed hypothesis is based on the assumption that graves generally represent the distribution and the intensity of settlement, both in time and space, and that grave construction (variation and monumentality, etc. ) is in some way representative of the structure of the Iron Age society (Hyenstrand 1984(Hyenstrand :212, 1996).
Other attempts have also been made to isolate regions or societies on a more comprehensive level (e.g., Ramqvist 1991).Per Ramqvist has, for example, suggested that the socio-political structure in Scandinavia, outside Denmark and Scania, was made up of thirteen petty kingdoms, all more or less independent but connected with each other in different ways (Fig. 1).With reference to the Roman author Procopius (ca.500-560 AD), Ramqvist argues that "the number of areas archaeologically corre- sponds exactly with what Procopius tells us about Thule, namely that this part of the ancient world was inhabited by thirteen more densely populated tribes, each led by a king" (Ramqvist 1991:306).Åke Hyenstrand has outlined a similar hypothesis, namely that archaeological regions correspond very well with the distribution of the Iron Age Nordic tribes as suggested by the author Jordanes (1996:132).These suggestions are problematic because the information that Roman authors provide about their northern neighbours in the form of written documents is all outsiders' accounts, and they need to be evaluated with great critical consideration rather than be taken for granted (Wells 2001:15).On the basis of references to Greek and Roman authors, many modern archaeologists and researchers draw maps that show the locations of different groups, kingdoms and regions, situated in specific places in Europe.Despite the problems of using the information that the classical texts provide, the archaeological correspondence with such a division is also a matter for critical evaluation.
There is no doubt that the application of the region-concept to prehistoric material is fraught with several methodological diAiculties related to general clas- sification.There is no single standard method for defining a clear-cut homogeneous region because the concept also concerns aspects of shared knowledge (culture), social interaction and differentiated relations among people within a geographical area.This will result in material complexity rather than a homogeneous pattern.It is also important to keep in mind that the region-concept is an abstraction and nowadays usually an instrument for the national or local planning ad- ministration.There are no naturally fixed regions out there, ready to be classified and determined (e.g., Helmfrid 1969).The principal objec- tion against using general spatial division models is that they usually are constructed in accordance with modern administration purposes.But we cannot take for granted that the ancient experience of spatial dimen- sions was limited or ordered in the same way as in modern social sit-  uations (Andrén 1987:23).Anders   Andrén argues that extra caution is needed when using clear-cut geo- graphical division models in a dis- cussion of the spatial dimensions of ancient territories, because we simply cannot assume that we are dealing Fig. 2.MapshowingthefrequencyoflronAgegravesin with administrated homogeneous northern Swedett (after Sveriges Nationalatlas 1994.33) units.
Hyenstrand's region-division hypothesis is based on one-factor analysis, namely the grave material.The prehistoric grave material is perhaps the type of information, both quantitatively and qualitatively, of which we have the best overall understanding, but the problem is that we cannot assume that the grave material is co-ordinated with one or several other more important manifestations in a society.Manifestations of similar grave types, for example, may represent quite different social forms, both during the same period as well as in different periods of time (e.g., Hyenstrand 1984:211).The application of the archaeological division model is problematic in this respect because it becomes limited to those areas where visible graves appear.In accordance with Hyenstrand, archaeological regions are recognized by the distribution of similar grave types, and this clearly makes his region approach an issue of homogeneity.In northern Sweden, Norrland, visible graves are less frequently distributed compared to the southern parts and therefore not further discussed in Hyenstrand's survey (Hyen- strand 1984:212f).The limited number of visible Iron Age graves in Norrland are, however, of the same types as the graves that occur elsewhere in Sweden (Selinge  1976; Sveriges Nationalatlas 1994).Thus, it might be of some relevance to take a closer look at the spatial significance of the northern grave tradition.
The main purpose of this paper is to discuss the distribution and significance of Iron Age graves in the interior of central Norrland (Fig. 2), which comprises the provinces of Härjedalen, Jämtland, and the western parts of Medelpad and Angermanland.I will further relate this discussion to the issue of how an archaeological region should be conceived of, and discuss the problems of applying the concept to Iron Age burials in the interior of central Norrland.But first I will give a short introduction to the grave material.
Ctv. rent Sw:edishArchaeologgi, Voh l0, 2002  IRON AGE GRAVES IN THE INTERIOR OF NORTHERN SWEDEN The Iron Age burials, settlements, finds and depositions in northern Sweden have been widely discussed over the years in the archaeological literature.This is, however, not the right place to present a general research history (see Baudou 1992; Burenhult 1999; Selinge 1976; Zachrisson 1997 for further references).I will here focus on the discussion of how different burial monuments are related to the presumed division between the prehistoric agrarian settlement areas (Sw.bondebygden) and the prehistoric hunting grounds (Sw.fängstmarken) in northern Sweden (e.g.Baudou 1977; Selinge  1976, 1977; Sundström 1997; Sveriges Nationalatlas 1994; Zachrisson et a1.1997).
What is generally considered as agrarian settlement areas are those areas where we later find medieval villages documented as sedentary agrarian in the historical and archaeological records.The prehistoric agrarian settlement areas have in a simplified way often been classified by their location to the best areas for agricultural activities, with respect to soil type, fertility, etc.There are large parts of northern Sweden, especially in the interior of central Norrland, which are less suitable for agricultural activities, and these have therefore been classified as hunting grounds (e.g., Selinge  1976, 1979).
These two economically described areas focus primarily on the topographical prerequisites for different subsistence activities, and it is against this background that the archaeological materials in central Norrland to a large extent have been discussed.
The interesting issue of interpretation here is, however, how the archaeological remains, in this context the grave material, correspond to these classified areas.In the discussions about the representative meaning of the Iron Age graves in northern Sweden it is generally held that mounds of various size, usually a central cairn or stone-setting covered with earth, are connected with the burial of people with an "agrarian back- ground" (Baudou 1992; Liedgren 1988; Selinge 1976).This interpretation can be questioned since a direct connection between burial form and a specific mode of subsistence is problematic and can therefore not be generalized.Lars Liedgren (1988) has suggested that the tradition of building mounds symbolically marks the transition to a more sedentary agrarian life along the Norrlandic coast during the first centuries AD.This is in my opinion an interesting thought, but one can also contemplate the symbolic significance of the cairns and stone-settings that exist contemporaneously and sometimes in the same areas as the mounds (Lindberg 1999).Do they represent another form of subsistence economy or are they perhaps significant to groups with different ethnic or cultural orientation in the society?Ifthat is the case, how is such a prehistoric society constituted?The situation along the Norrlandic coast has indeed an important role when discussing the spatial distribution of mounds and stone-settings in the inland parts of central Norrland, because it puts focus on the representative meaning of the different forms of burial constructions and traditions.
As shown in figure 2, there are two types of Iron Age graves distributed in the interior of central Norrland -stone-settings and mounds.The stone-settings, which are sometimes covered with a thin layer ofturf or sod, can vary in form and are generally Current Swedish Arehaeology, Voh l0, 2002 located close to lakeshores and watercourses.These stone-settings are classified according to their location as lake graves (Sw.insjögravar).A closer look at the distribution of the lake graves in the districts of Härjedalen and Jämtland (Fig. 2) shows that they occur along main river courses within the inland water system.The lake graves have, according to grave finds and "C-analysis, been dated to the entire Iron Age period, from the last centuries before Christ to 1200 AD.There are approx- imately 230 lake graves recorded at 64 sites within the Härjedalen/Jämtland area (Selinge 1976:40;Zachrisson 1997:22ff).
Mounds with an earth fill, which comprise the other grave type, are more frequently distributed within the Härjedalen/Jämtland area (Fig. 3).The number of grave mounds is approximately 580, distributed on more than 330 sites (Selinge 1976:49-54).In northern Sweden the tradition of building mounds existed from the 2"4 or 3"4 century AD (with reference to dated grave mounds in Trendelag in Norway and in the coastal area of Medelpad) to the end of the 11'" century AD (Baudou 1992; Selinge 1976:47; Zachrisson et al. 1997).Although very few mounds in the interior have been excavated and dated, it is plausible to consider the mound-building tradition in the interior of Norrland as chronologically parallel to mounds in Trendelag and the Medelpad area, that is, approximately from the 2ed or 3rd century AD and onwards.We may therefore assume that both stone-settings and burial mounds with an earth fill in the Jamtland interior of Norrland appear parallel with the graves along the Norrlandic coast and in Trkindelag during the Early Iron Age (e.g., Baudou 1992: 115ff;Selinge 1976:47).
We can conclude from the distri- bution ofthe lake graves and mounds, that there are several concentrations of graves to deal with when inter- preting the Iron Age burial tradition in the interior.One large concentrae tion is located in the Storsjön lake ~, . . .Lake gravas district in Jämtland.It is here that the majority of lake graves and burial Hatledalen Grave meunds mounds are found, sometimes localised in close connection to each other, as evidenced by the Rödön/Frösön area.It is interesting to note that Fig. 3.The distribution oflake graves and burial mounds burial mounds with an earth fill do in the provinces of Häjredalen and Jämtland. (After not only occur in the so called Se/inge)976.41, 49) Current Srvedish ArchaeotogJ, Vot.l0, 2002  agrarian settlement areas, such as the Storsjön lake district in the central part ofJämtland.
Approximatically 60 mounds also occur in the foothill area in western Härjedalen and Jämtland.Some of the burial mounds are located very high up in the tree-border zone at levels over 900 metres above sea level (Selinge 1976:55).These environments can, of course, not be connected with an agricultural economy except perhaps with regard to herding.In these areas we generally f1nd stone-settings but also a limited number of burial mounds along lakeshores, for example around the lakes Annsjön and Singsjön in Jämtland, and in the mountain foothill valleys, such as in the Ljungdalen area in Härjedalen.The stone-settings are furthermore not only limited to watercourses but can also be found in the boundless hill terrain between 600 and 800 metres above sea level.Four of these graves have been excavated and are dated to the Late Iron Age, ca. 800-1000AD (Hansson 1997).
How should the variability and sporadic distribution oflake graves and burial mounds be understood and interpreted in terms of cultural tradition and regionality?Before turning to this issue I will discuss some theoretical points of departure.

CULTURAL BLINDNESS AND MENTAL MAPS
Scholars of anthropology have often pointed out that people are usually blind to the meaning of cultural phenomena and symbols in their home environment (e.g., Ehn & Löfgren 2000:12ff; Hirsch & 0 'Hanlon 1995; Tuan 1977).According to the anthro- pologist Abner Cohen: "Symbols are essential for the development and maintenance of social order.To do their job eff1ciently their social functions must remain largely unconscious and unintended by the actors.Once these functions become known to the actors, the symbols lose a great deal of their efficacy" (Cohen 1974:8).We sometimes see symbols actively used in advertisements and political propaganda, but the power of symbols is also silently present in the material world, indirectly structuring social relations in people's daily life.Material symbols are in this sense essential for the maintenance of social order, and they do not necessarily always function as overt signals or political manifestations.The symbolic meaning of things is, as Cohen and others maintain, hard to grasp because the meaning also tends to be fluid and may change depending on the contextual situation.We usually cannot explain the signif1cant meaning of the symbols and actions around us because it is "tacit knowledge" to us.This does not imply that the meaning of cultural expressions around us in the home environment is less essential to us.This kind of tacit knowledge is of great importance to people because it makes them feel safe and at home in their part of the world.Not only burial monuments, buildings, roads, etc. , but also topographical contours and contrasts in the landscape have cultural signif1cance in this respect.According to this line of argument, places and physical phenomena in the local community can be described as check-points on the mental map-sheet in people's minds (Ehn & Löfgren 2000:52f).One may simplify this description and say that these mental map-sheets to a large extent are the core of cultural cognition.In a general sense they activate our moral resources to judge right from wrong and guide us in various social situations.the assumption that the Iron Age world-view in the interior of central Norrland was based on such mental maps.The distribution of lake graves and burial mounds was one ofmany check-points to refer to.There were, ofcourse, additional material features of significance in the interior which we are less able to determine, such as remains from settlements and buildings.
The tradition of building lake graves and burial mounds throughout the Iron Age period was not a homogeneous project because ancient life was not administrated in such a way.The grave-building tradition can, of course, be viewed as the result of people from outside areas moving in as settlers, but variation in burial form may also have been determined by tradition and expectations, which continuously created attachment to the homeland among people within the community.According to the geographer Yi-Fu Tuan (1977), attachment to the homeland is a common human emotion, although its strength varies among different cultures and historical periods.
A homeland has its landmarks, which may be features of high visibility and public significance, such as burial monuments, cemeteries and ritual sites (Tuan 1977:158f).These visible signs serve to enhance a people's sense of identity; they encourage awareness of and loyalty to place.But attachment of a deep though subconscious sort may also come simply with familiarity and ease, with the assurance of nurture and security, with the memory of sounds and smells, of communal activities and domestic pleasures accumulated over time.It is important to state that attachments of this type can form without the memory of fear and superiority vis-a-vis other people (ibid).
The people living in the interior of central Norrland were probably familiar with the burial monuments as they wandered along tracks and paths in the landscape, and the presence of ancestors in the ground was no doubt explicit knowledge to them.The more abstruse significance of the various forms ofburial monuments, which is difficult to appraise, lies in how these graves structured the minds of people passing by or paying visits.One may, at any rate, consider the visible Iron Age graves as checkpoints in people's world-view, and this is central to our understanding of what these monuments signified.This line of argument leads us deeper into the issue of how the archaeological region-concept can be applied to the Iron Age grave material in the interior.

REGIONS AS IMAGINED COMMUNITIES
From a general point of view, a region can be considered as a sort of over-bridging network of communities, in which people exchange knowledge, material necessities and artefacts with each other in various forms of interaction.It is also possible to describe a region as a hierarchy of spatial "us and them" relations where the social situation determines the range of interactions, affinity and group identity.It is us here in this family as opposed to the family in the neighbouring house, it is us here in the southern part of the village as opposed to them in the north, it is us living in this parish as opposed to people from the outside world, etc.A region is therefore not a fixed and clear-cut geographical area.It cannot be characterised directly by the distribution of similar artefacts or grave types or other similar features.A region can instead be Current Sivedish Archaeology, Vok l0, 2002 described as a normative view of the world, which people share with each other within a relatively wide geographical area.On the individual level people may, of course, have different viewpoints and horizons, which may cause overlaps and fluidity, but the essential point is a general sense of community.One may complement this description and consider a region to be what the historian Benedict Anderson calls an imagined community (Anderson 1983).All sorts of communities are imagined but not always in the same way.A community is, according to Anderson, imagined because the members of even the smallest social formation larger than primordial villages of face-to-face contact (and perhaps even these), will never know most of their fellow members, meet them, or even hear ofthem, yet in the minds of each lives the image oftheir communion (Anderson 1983:15).
As can be seen in figure 3, there is no use in trying to define a homogeneous region or a unified and fixed territory on the basis of the distribution of burial mounds or lake graves.Both lake graves (stone-settings) and burial mounds occur in a variety of topographical environments, and sometimes together in several areas in Norrland.We are therefore not, as Hyenstrand suggests, primarily looking for homogeneity in the archaeological material (Hyenstrand 1984:211).Since we are discussing the distribution pattern ofmaterial culture, our analysis should not come down to the basic configuration of the concept of culture, that of an overarching orders of orders from which more limited configurations can be deducted (e.g., Geertz 1993).We must instead direct our attention to how people within a settled area may have interacted and to the variety of ways these people manifested their culture.
In line with Clifford Geertz' concept of culture, a cultural region can be described as a less tightly coherent but nonetheless ordered system, made up ofclusters of clusters of significant symbols (Geertz 1993:408).Since we are dealing with a long-term perspective, it is useless to expect a static geographic setting.Cultures and regions are human products, and since it can be expected that people continually move in and out of areas, any homogeneous pattern of culture will eventually dissolve.The problem is as much a matter of determining independencies as interconnections, gulfs as well as bridges.Following Geertz' image of culture, one can compare a region to the image of an octopus, an octopoidal system "moving" by disjointed movements of its tentacles (ibid).Since we will find clusters of graves of different forms occurring in various parts of the interior of central Norrland, it is problematic, and perhaps also illusive, to search for a holistic and coherent region.Perhaps a more appropriate approach is to focus on the tentacles.
I will illuminate this assumption by referring to two areas that have a relatively high concentration of visible graves.One is the Storsjön lake district in the central part of Jämtland and the other is the Funäsdalen/Ljungdalen area in the northwestern part of Härjedalen (Fig. 3).Both the mounds and lake graves are distributed within reach ofeach other in these areas (e.g., Magnusson 1989;Sundström 1989).In the Funäsdalen/ Ljungdalen mountain area there are several mounds located on very high levels in the mountain hill terrain (Fig. 4).
The area also exhibits a large number of hunting pits, settlements sites and burial Curreni Swedish Archaeology, Voh 10, 2002 Fig. 4. Vievv of the Tänndalen/Funäsdalenllj undalen area in the northivestern part ofHäjredalen.
(Modified after Sundström l989: 88) grounds, such as Krankmårtenhögen and Vivallen, dating to both the Early and the Late Iron Age (Sundström 1989;Zachrisson 1997).The distribution of stone-settings, mounds and inhumation burials (without visible construction) within the Funäsdalen/ Ljungdalen area shows that the people living there not only altered their burial practices during the course of the Iron Age, but they were also living in an 'open society'; it was hardly the conservative isolation that we sometimes assume, but instead continuously open to external contact and influences.As far as the Iron Age grave material is concerned, ritual life and culture in the interior of central Norrland seem to have been structured and differentiated in a variety of ways, and this also seems to be the case among the Saami groups later on according to the ethnographic record (Manker 1961).
Although we may assume that there existed a hierarchy of "us and them" relations between groups living in the interior of Norrland, it is less fruitful to argue that these relations were homogeneously expressed in the grave material.These kinds of"us and them" relations can instead be assumed to be flexible and activated in various ways over time, mainly depending on the situation and on various forms of power struggle within the local community.

BEYOND REGIONS
The limited number of visible graves is far from representative ofthe overall population in the interior of central Norrland, and since we lack knowledge of alternative burial forms it is problematic to associate the stone-settings and mounds with the burial practice of the whole population.However, we should not be constrained by what we CurrentSrvedishttrchaeotogy, Fig. 5.The folklandsin the kingdom ofcentralNorrlandin the Migration period according to Per HRamqvist (I99l. 306).
consider to be limitations of our data in our interpretations, because it frequently sets up a vicious circle (Bender 1978).ln this case there is enough material to exclude any sug- gestion that the graves repre- sent some kind of incidental occurrence, and we definitely have to put more meaning than that into our interpretations.Semi-sedentary groups, with which we are likely dealing here, cannot be autonomous.They are dependent on contact with other neighbouring groups and have a need to move across the landscape to get ready access to different resources (e.g., Bender  1978:210).We may very well expect that there existed such dependencies and reciprocal relations among the people living in the interior of Norrland, but we don't need to presume that reciprocal relations always were significant to groups living within fixed territories.
One of the thirteen petty kingdoms outside the Scania and Denmark area during the Migration period was, according to Ramqvist's suggestion, located in central Norrland (Fig. 5).Ramqvist concludes that there are great possibilities for isolating different kingdoms within Scandinavia during the Migration period.The central Norrlandic coast and the Storsjön lake district constitutes, according to Ramqvist, one such kingdom (1991:306).But what is a petty kingdom, and what limiting factors are significant to its geographical relations to other kingdoms and territories?I must admit that I am not able to follow Ramqvist's conclusion, mainly because there is no empirical discussion about the isolated character of these kingdoms.When focusing on similarities and differences in the archaeological record, it is here perhaps of some relevance to bring the distribution of asbestos-tempered bucket-shaped pottery, dating to the Roman Iron Age and the Migration period, into the discussion.Ramqvist (1992)has noted that this type of bucket-shaped pottery is common in various parts of Norway and northern Bohuslän, but occurs also at settlement sites and in graves in central Norrland (Ramqvist 1992:186).At Backnorholmen in Jämtland, for example, a bucket-shaped pot was found in a grave dating to the Migration period (Sundström 1989).In relation to SW Norway, which has most of the bucket-shaped pots, Trendelag on the one hand and northern Bohuslän on the other can be compared with central Norrland.Asbestos-Current Swedish Archaeotogy, Vot.l 0, 2002 tempered pottery of various types may additionally be seen as one of the few cultural characteristics which exhibit continuity in central Norrland, &om the Bronze Age/pre- Roman Iron Age and further on into the Migration period (Ramqvist 1992:186);and in relation to the region-concept, discussed as a cultural phenomenon in the previous section, it is perhaps relevant to regard the wide but sporadic distribution of bucket- shaped pottery as one of the many interrelated tentacles of culture and its movement over time.
Returning to the discussion of burials, the variable distribution of different grave forms in the interior of central Norrland indicates that people in fact were frequently moving from one area to another, which may have contributed to maintaining reciprocal relations between near and distant groups.The various groups living in central Norrland were probably not at all limited to spatially f1xed territories.A few groups may have pursued a rather sedentary way of life, but certainly not the majority ofthe population.
This assumption opens up for a scenario where several burial traditions were practised and perhaps altered, as people regularly were moving between different areas in the interior.
It is not just the distribution of grave forms which makes the discussion of ancient regions complicated.There are also variations in the find material.The coastal area and the inland part of central Norrland have to a large extent similar burial practices, although there are interesting differences concerning the deposition of animal bones in some of the graves.We know that Iron Age graves in Scandinavia usually contain a large amount of animal bones deposited among the cremated human bones, and this is also the case in central Norrland (Iregren 1997(Iregren , 1989;;Zachrisson et al. 1997).Several lake graves in the interior of Norrland contain not only cremated animal and human bones but also depositions of antlers of elk and reindeer.In the Early Iron Age cemetery Kran/rmårtenhögen in northwest Härjedalen, several graves contained the remains of antlers which had been deposited on different occasions (Iregren 1989; Zachrisson et   al. 1997).The distribution of excavated graves containing antlers of elk or reindeer is presented in figure 6.
It is interesting to note that antlers not only occur in lake graves such as stonesettings located near watercourses.In one of the burial mounds (no.I) on Långön in Ångermanland, antlers of reindeer were found in the ftll (Arne 1926:87).This shows that the tradition of depositing antlers in the graves was not only limited to lake graves located in the backwoods and mountain areas in the western part of central Norrland.
The Långön/Hoting area, located approximately 150 kilometres away from the Lake Storsjö area in the northeastern part of Jämtland, may perhaps represent a community of its own, but it was not isolated from the practice of depositing antlers on graves as we know it from the inland area of western Härjedalen and Lappland.

CONCLUSION
It has often been suggested that the tradition ofbuilding lake graves and burial mounds in the western part of Jämtland/Härjedalen can be considered as the result ofcolonisation by traders, possibly from Svealand in central Sweden, from Trendelag, or from the Serning 1962, 1966).Selinge (1976)  assumes that the tradition of lake graves and, later on, the limited number of mounds in the inland hunting grounds, may have been initiated by people coming from the central agrarian areas, who were visiting the peripheral hunting areas for economic reasons.Burial was thus not the main reason for visiting the hunting grounds, according to Selinge.He further argues that a new burial tradition with new grave forms should likely be considered as a new religion, and this cannot have been an internal invention (Selinge 1976:69).
Why not, one may ask?This line of argument is just a detour to avoid a scenario where change is assumed to occur internally, without external input to the existing system.People may move to and fro between different areas and exchange things and knowledge with each other, and this is essential to human sociability and cultural variation (Carrithers 1992), but it is not reasonable to consider the limited number of graves as incidental burials of visitors/exploiters who came from outside, and were not living in the interior.Other archaeologists have explained the spatial differences between the central agrarian areas, where burial mounds dominate, and the hunting areas, where lake graves frequently occur, in terms of ethnic differences between indigenous Saami people and Nordic/Germanic groups colonising the interior (Zachrisson et al. 1997).This as- sumption is indeed an interesting approach, but it is also problematic, because what in reality is materially significant when it comes to the determination of ethnic symbols?Studies of the Iron Age grave material from Långön in central Norrland (e. g, Olofsson 2000) have, for example, shown that a strict classificatory division between ancient Saami people and Nordic/Germanic people not only is difficult to make, but it also serves to maintain a too simplistic and generalised picture of prehistoric culture.
In my opinion we cannot accept such a dualistic approach because it does not correspond to the material complexity in the archaeological record.Why only two contracting parties, for that matter?Why not a hundred groups speaking a dozen different tongues, devoted to a variety of subsistence strategies, but still interacting within areas conceived of as imagined communities?
Current Swedish Archaeotogt~, Vot.10,2002 As I have commented above, I am not inclined to associate any specific grave type with any particular ethnic group ofpeople because the complexity ofmaterial influences does not allow for a strict division between ethnic categories.Material expressions and symbols may occasionally be significant to both ethnic identity and economic strategies, but that is not a satisfying approach to the region issue.Instead we may assume that there existed a hierarchy of "us and them" relations between groups in which identities and power were expressed and negotiated, and we may also assume that people were continuously moving in and out of areas in the interior.This may explain the material complexity where objects, ideas and practices were exchanged, shared or discarded by groups.The people living in the areas presented above were probably very well accustomed (whether they liked it or not) to a world of material complexity and differences, and that is why regions do not need to be homogeneously specific in character.The region-concept represents one of many important aspects of cultural configuration, but in a long-term perspective culture moves like an octopus, which makes regional boundaries hard to grasp.English revised by Laura 14rang.

Fig
Fig. ).The Scandinavian petty kingdoms (regions) in the Migration period.(After Ramqvist l99l:306) An important point of departure in the following discussion of the region-concept is Current Swedish Archaeology, Voh l 0, 2002 Culiure Moves Like an Ociopus I 3

Fig. 6 .
Fig.6.Antlers and hanes of elk, reindeer deposited on or in Iron Age graves in central and northernSkeden.(After Zachrisson 1997.198)