An Iron Age Magnate Farm at Odarslöv – a local centre in the realm of Uppåkra

Bertil Helgesson & Håkan Aspeborg 2017. An Iron Age Magnate Farm at Odarslöv – a local centre in the realm of Uppåkra In this article, a newly excavated Iron Age farm with an associated gravefield and its surrounding landscape outside the city of Lund is examined, discussed, analysed and placed in its context. The biography of the farm spans about 400 years, from the Late Roman Iron Age to the Early Vendel Period. Through all phases, the farm remains on a restricted plot and shows both stability and prosperity. It is argued that the farm, based on its size, the number and types of buildings and the find material, can be considered a magnate farm. As magnates, the farm owners must have played an important role in the local community during a time when society underwent change and transition. The farm and the farm owners’ relation to Uppåkra and the rulers of the central place is discussed.


Introduction
"An international collaboration of scientists opens new horizons in powder diffraction and expands the potential for determining the structures of new and more complex materials" (ESS).The European Spallation Source, ESS AB, will be at the forefront of research when it is commissioned a few years from now.The northeastern outskirts of Lund, Scania in Sweden, are being transformed into a modern research area.Just a few years ago, these areas were farming land, and had been so for thousands of years.The archaeological investigations that preceded ESS have shown that the area was densely inhabited during the Iron Age.This paper focuses on the period around AD 250-650, and analyses one of the farms from the area.The results of the archaeological investigations at ESS described in this paper are based on the report from the investigation (Helgesson & Lindberg, 2017).To avoid confusion, the term ESS concerns the entire investigation area, and Odarslöv 51 denotes the Iron Age magnate farm.

Searching for the past
Even before the archaeological investigations were performed, problems relating to ESS were formulated.A central problem is the location of ESS on the northern-most point of the Romele ridge, and its relationship with the surrounding areas.Other problems concern its buildings and farming history in a long-term perspective, as well as communications and meetings.
This paper deals with the Iron Age settlement from ESS, i.e. the Iron Age farm (Odarslöv 51), the burial ground, the social, economic and ritual context, and the geographical location in the borderland between the plain of Lund and the Romele ridge area.The position of ESS in the landscape just 10 kilometres from the central place of Uppåkra also raises the question of the relation between the two sites (Fig. 1).
In settlement archaeology, we seldom have the chance to excavate more than a small part of a farm or a village/hamlet.Activity areas at a distance from the buildings and gravefields connected to the settlements are rarely found and investigated.In the present project, we had the opportunity to remove the topsoil from an extensive area.In total, the trenches in all three settlement areas of the ESS project covered 103,000 m 2 .In spite of that, and as often happens in contract archaeology, some parts of the farm Odarslöv 51 lay outside the development site and could thus not be excavated.

The Iron Age settlement
The ESS area is situated in the hummocky landscape northeast of the city of Lund.The area has been more or less intensively settled from the Neolithic until most recent times.Remains from Iron Age settlements were found all over the area but with a distinct concentration in its southern-most part.Here a farm was established in the mid-third century AD and remained at the same location for around 400 years.
The farm is situated on a sandy plateau.The area is dominated by arable land, and this corresponds to the situation in older times according to a map from the 1790s (LMV L271-6:1).To the south, the farm faced towards meadows and wetlands, and to the north is a marked height.Access to fields and grazing might have played a role when the farm was established.The situation next to an ancient communication route, which survives in the Odarslöv road, might also have contributed to the location.The road leads north up to Gårdstånga, which was an early medieval royal estate.Excavations have dated the estate at least to the Vendel Period/Viking Age (Söderberg 1995).

The farm
During the period 250-650 AD, i.e. the Late Roman Iron Age to the Early Vendel Period, according to Swedish terminology, four farms (four generations of the same farm) succeeded each other at Odarslöv 51.Twelve excavated house foundations and some other structures situated in an area measuring 100 × 50 m were interpreted as belonging to the four generations of farms.The farms have been dated by radiocarbon analysis as well as by typology.When separating the different farms from each other, the geographical distribution of the houses has also been used.
A Bayesian approach has been used in which the calibrated 14 C results from the site have been subject to Bayesian statistics by Bettina Schulz Paulsson (2015).Our dating of the phases before the Bayesian modelling is presented within brackets.A total of 32 radiocarbon results were used for the Bayesian chronological model.According to the Bayesian statistical method the timespan of the farm could be narrowed down to 235-615 AD.
PHASE 1, BAYESIAN MODEL ABOUT 235-335 AD (95.4%) (250-325 AD) The earliest farm was established in the mid-3rd century and consists of two houses: a main building and a minor hall (Fig. 2).The main building, house 105, is about 33 m long and more than 5.5 m wide.Judging by the positions of the trestles, the house had several sections corresponding to different functions or rooms.There is a clear division with an entrance or passage in the middle, and small rooms at the short ends.The roof bearing posts are placed with a wide interval in the centres of both the east and west sides, suggesting two larger rooms.One interpretation is that the eastern part of the building is a barn and the western part a residential area, where a storage pit with carbonized grains may have been a pantry.An alternative interpretation is that the western part of the building was a storage room and the eastern part was the residential area.The first interpretation is perhaps most likely.In a contemporary house with much the same layout as house 105, at Örja house 36, the kitchen and dwelling were situated in the room furthest to the east (Schmidt-Sabo 2013: 281-283).Some 30m to the east of the main building, a minor hall, house 116, was located.It is a small building, built on a small sandy elevation on its own.The house is only about 8 m long and about 5 m wide, with bent walls.There seems to have been an entrance in the southeastern part of the building and perhaps another in the western gable.This small building has a parallel in the well-known ceremonial building from Uppåkra (Larsson & Lenntorp 2004).It is smaller than the Uppåkra house, but the two houses share the same proportions.There are also some other differences.The find material from Uppåkra is rich and indicates various aspects of religious and social life of the Iron Age (Larsson (ed.) 2004).There is no find material from house 116 and there might be several reasons for this.The area was heavily ploughed over, and organic and non-metallic finds might have been destroyed.We know for certain that bones are badly preserved at ESS.The area was scanned with a metal detector but no metal finds were recovered.The most logical interpretation is that depositions in and around house 116 were not preserved, and that more prestigious objects were moved to its successor during phase 2. House 116 is contemporary with the Uppåkra house or slightly younger.It is most likely that house 116 was built with the Uppåkra house as a model.Therefore, these two houses must be seen in the same social and ritual context (see below).House 36 from Stora Hammar (Björhem, Fabech & Jönsson 2011: 92) is almost identical to house 116, including its size.It could be dated to 340-540 AD and is slightly later than house 116.It is situated close to a concentration of houses from the Late Roman Iron Age and the Migration Period and might have served as a ritual building for the community.This indicates that house 116 at Odarslöv 51 does not stand alone, but that different communities had their own ritual buildings for local needs.
A row of five hearths to the west of the farm perhaps marks its western limit.Their function is unclear, but one interpretation is that they served as light or heat sources.Another possibility is that they had ritual or other functions in connection with the establishment of the farm.
An area with ovens and other features that can be related to food preparation was also established at this time (see below) some 30 m to the northwest of house 105.The area has been interpreted as kitchen, bakery and perhaps brewery.
The burial ground is contemporary with the first farm, which was established about 100 m to the east (see below).PHASE 2, BAYESIAN MODEL ABOUT AD 288-393 (95.4%)This phase is characterized by a new main building and a new hall (Fig. 3).A new and larger main building, house 107, was erected directly to the south of house 105.The house is 40 m long and 6m wide, and had some kind of superstructure, probably a loft.This indicates a greater need for space, probably associated with increased yield and surplus production.The house has three or four entrances, perhaps even more, and one with some sort of vestibule.A kitchen dwelling room with a hearth lay east of the central entrances.Apart from that, no room division was observed, but a varying distance between the roof-bearing posts indicates a functional division.Actually, the sectioning of the house much resembles that of its predecessor house 105.House 107 has several parallels in southern Sweden and Denmark.These can be exemplified by a house from Ringsted on Zealand, which can be dated to around the beginning of the common era (Poulsen 2008: 3;Østergaard Jensen 2008: 10 f ), House 3 from area 13 of Mellanbyn, Malmö (Friman, 2008: 96), and House XXXII from Snöstorp, Halland (Carlie 1992: 18).
During this period, house 116 was replaced by a much larger hall, house 109.The hall was relocated to the western part of the farm area and much closer the main building.House 109 is about 23 m long and about 5.5m wide.
It has concave gables and two posts close to the north wall might be traces of a high settle.House 109 (and also the later hall, house 108, see below) shows outer similarities to the hall in Forsand in Rogaland (House II), Norway (Løken 1983) and is reminiscent of the hall in Østergård in Jutland (Ethelberg 2003: 312-314).The basic plan also closely resembles hall buildings from the Roman Iron Age and the Migration Period in Gudme (Østergaard Sørensen 1994: 28-33).Concave gables are relatively uncommon construction details, as e.g.found in the Vendel Period/Viking Age hall buildings in Järrestad (Söderberg 2003: 128-133;Söderberg 2005: 192-196).Thus, house 109 has parallels in a number of high-status milieus in the Nordic countries.
There is continuity from farm 1 to farm 2, indicated by several facts.House 107 and the older house 105 lie parallel and very close to each other.They are almost of the same size and are sectioned in roughly the same way.The burial ground, as well as the area with ovens and other structures that can be related to food preparation, was also continuously used.
A discontinuity may be seen in the new location of the hall at this time.This could suggest an increased need to control the functions of the hall.The social and ritual functions thus became a more integrated part of the farm.According to Herschend, the hall also became the residence of the farm owner, or the magnate, and his family in the middle of the first millennium (cf.Herschend 2009: 259-260).If this was the case at ESS, the main building (house 107) most certainly should be related to economic functions, and dwellings for farmhands, servants, slaves, etc.One may also speculate about whether it is the hall's character as residence and thus as the public building of the farm that is the reason it was moved towards the front of the road, where it could be admired by passers-by.PHASE 3, BAYESIAN MODEL ABOUT AD 356-515 (95.4%) (AD 400-550) During the Migration Period, the settlement expansion peaked and the farm then comprised six buildings (Fig. 4).A new main building, house 106, was constructed and given a slightly different direction compared to the main buildings during the earlier phases.The house is more than 39m long and about 6.5m wide.It has an entrance on the south side and likely a loft, similar to house 107.A relatively large variation in the width of the roof-bearing trestles as well as the span length indicate that the house was multifunctional.There are also traces of stalls, indicating that the eastern part of the house was a byre.It is a typical South Scandinavian house with a central entrance room and has parallels in house VI at Brogård, Halland (Carlie 1992: 25, 27) among others.
The hall, house 109 from farm 2, was replaced by a similar building, house 108, which almost completely overlapped the older one.Here too there are two posts close to the northern wall, indicating a high settle.Perhaps the parts of a vessel that was found at the building's west end and a grindstone from one of the postholes for the high settle are house sacrifices, emphasizing the building's ritual function.Parallels to this building have been discussed above.
The 17-m-long house 103 can also be attributed to this phase and is interpreted as a barn, or an additional multifunctional building.This type of building with clearly grouped roof-bearing posts and longer spans has several parallels in Scania and Halland (cf. Carlie 1992: 24, 26;Tesch 1993: 43;Björhem & Magnusson Staaf 2006: 98-99).Often the entrance is directly to the west of the centrally grouped roof-bearing posts.A four-post barn, house 101, may also be added to this phase.The house is unusual with a fifth pillar, a plank, on the north side.This construction detail has a possible parallel in Ringsted on Zealand, where at least two four-post barns have a similar feature (Christensen 2011: 28-29, 132, 157).The house may have been used for storing hay or grain.About 20m to the north of the main building, house 104 was built close to the area with ovens.This special building was probably built for the processing of grain.Close to house 104, house 114 is situated.This house might have been used for storage.A row of pits on the western side of the farm is probably a boundary marker.
The burial ground was still in use at the beginning of this period, but then burials ceased.Where the people from Odarslöv 51 continued to bury their dead is not known.
During this phase, there is a remarkable change in the roof-bearing posts.Preserved post colourings show that they evolved from round logs, in farms 1 and 2, into plank-shaped poles, probably hewn by a broad-axe.At Odarslöv 51, this was evident in the main building and the hall, while other buildings seem to have made use of rounded or slightly oval posts.Plank-shaped pole colourings with similar shape and dimensions occur in a variety houses from the middle and Late Iron Age, and especially in buildings that can be attributed to the elite (Karlsson, in prep.).PHASE 4, BAYESIAN MODEL ABOUT AD 440-615 (95.4%)ABOUT  In the Early Vendel Period, the last farm was established at Odarslöv 51.This farm only consisted of a main building, house 102, and a four-post barn, house 123 (Fig. 5).House 102 shows kinship with house 106, and the two houses lie in parallel.House 102 also shows similarities to the Vendel Period houses such as B31: III from the Köpinge area and house 1 at Uppåkra 29 (Tesch 1993: 76, 192;Runcis 1998: 17-19).The farm's scope has now been drastically reduced, compared with farm 3.No fences could be identified belonging to this farm.
This period can be seen as a decline, when only the main building remains.Still, house 102 has a considerable length of about 40m.The location of house 102 rules out that hall 108 was still in use during this period.This may indicate that the farm owner had lost his predecessor's status.It might also be interpreted as that he was still in control of the land but living elsewhere on another farm (cf.Herschend 2015).There is a tendency that the main buildings, as well as the halls, successively move from the northeast toward the southwest from phase 1 onwards.There is always a possibility that a fourth hall, connected with house 102, lies beneath the Odarslöv road or further south, and outside the excavation area.

The burial ground
On a slope overlooking the farm was a small burial ground.It was placed about 100m northeast of the farm and must have been visible from the road (Fig. 6-7).The burial ground consisted of twelve graves, and another five probable graves.All the graves were inhumation graves, orientated in a north-south direction.Only a few graves contained preserved bone material in the form of teeth or tooth enamel.Owing to bad preservation conditions, only a few individuals could be identified by age, while none could be determined as to sex.
Pottery as well as glass and amber beads are the dominant object categories found in the graves.Two graves stand out as particularly rich.Grave 77999 was furnished with two gold objects -a small disc and a ring too small to be a finger ring -and three pots (Fig. 8).Grave 79182 was furnished with a badly preserved fibula, a knife, 551 beads of glass or amber, and two pots.There were also three stone-paved pits at the burial ground, most likely foundations for standing stones or wooden poles.This indicates that the burial ground was marked in the landscape.
Apart from the ceramic vessels, the artefacts in the graves seem to belong to the dress, personal jewellery or small personal belongings such as the knife.
Owing to the almost total absence of fibulae, which is likely due to poor preservation conditions, the graves have been dated by bead typology, ceramics and radiocarbon analyses.Looking exclusively at the radiocarbon analyses, four out of five graves could be dated to between AD 180 and 420 (Cal 2σ).According to the Bayesian modelling, the timespan is between AD 217 and 339D (Cal 2σ) (Schulz Paulsson 2015).Grave 79876 was dated to period B1/B2 according to the radiocarbon analysis and to C1b-C3/D according to ceramics.Since there is uncertainty about the radiocarbon-dated material, i.e. charred oak from a coffin lid, the second dating is most credible.The burial ground was most certainly used during the Late Roman Iron Age and the Early Migration Period.This means that the burial ground could be connected with phase 1, phase 2 and the beginning of phase 3.Where people were buried later on is not known.
Late Roman Iron Age graves in western Scania seldom contain much grave goods, and even more seldom weapons, imports, silver or gold (Nicklasson 1997: 102-10;Björk 2005: 124-128).This is particularly clear concerning the graves around Uppåkra, where the equipment of the graves from the Late Roman Iron Age is modest, with a few exceptions such as a grave at Djurslöv, a grave at Önsvala and two at Kristineberg (Carlie 2005: 442;Branca 2001).This can be contrasted with southeastern Scania, for instance with the large gravefield at Simris (Stjernquist 1996: 200).It is therefore a difficult task to impose a ranking of social status of the graves in western Scania on the basis of the find material, as has been done in parts of Denmark (cf.Ethelberg 2000: 149-166).Gold is so rare in Scanian graves from the Early Iron Age that previously only 9 objects have been found, and grave 77999 from Odarslöv 51 must be considered as a relatively high-status grave (cf.Björk 2005: 126).Since graves and burial grounds are rare and small during the Late Roman Iron Age, one might look upon the grave itself as a status symbol in the area.To be buried singled out the family as elite, but on the other hand, among those families the furnishing of the grave was not an arena for social competition.Where the social structure and the elite are well established, there is no need to compete with burial display (cf.Halsall 1998: 148).
The burial customs in the Late Roman Iron Age in southwestern Scania and Zealand have much in common, which underlines shared religious and cultural concepts among the people in the area (cf.Björk 2008).The death rituals are in many ways elusive, owing to the burial customs and the poor preservation of both skeletal remains and metal.The small gold disc has the shape and size of a coin.It was found in the area where the mouth of the buried person is supposed to have been.The gold object might be interpreted as a Charon's obol.It has been proposed that this is one funeral custom that the Germanic people adopted from the Romans.This is obvious in present-day Germany and in Scandinavia as well, for example at the gravefield at Simris in eastern Scania.In many places, the coin has been replaced with a piece of metal, and in one of the Simris graves with a piece of amber (Stjernquist 1996: 201).Furthermore, the custom of Charon's obol seems to be common in rich graves in the Öresund area during Late Roman Iron Age (Lund Hansen & Rindel 2008).Lund Hansen & Rindel interpret the graves with Charon's obol as belonging to the elite of society.
The burial ground commemorates the owner's family at the farm.It is possible that rites honouring the forefathers took place here.

The burial ground
The close relationship between the farm and the graves at Odarslöv 51 is obvious and seldom seen in this period in southern Scandinavia.A question arises.Does the burial ground belong to the farm exclusively or was it shared by several farms?It is clear that there were contemporary farms in the northern area of ESS and also a single inhumation grave from the Late Roman Iron Age some 500 m to the northwest (Lagergren, Aspeborg & Artursson 2013: 28).This indicates a scattered settlement structure, where farms and graves were spread in the landscape.This has rather interesting parallels in other parts of Scania, where visible grave monuments from the Early Iron Age often lie scattered on uncultivable land and on small ridges surrounding the arable land.There is perhaps an idea behind the placing of graves.The monument was supposed to mark power, and property belonging to a farm or a village.This resembles the situation on Öland, where graves are sometimes located directly adjacent to the buildings, or outside the enclosed fields on the border to the outlying land, and sometimes lie in the border areas between villages, often along roads (Fallgren 2006: 117-121).The few graves at the burial ground show an imbalance between the number of people living on the farm and the people who were buried.Even if there are more undiscovered graves in the area belonging to the farm, this is not a unique case.Most likely the graves at the burial ground contain the members of the farm owner's family, probably those whom the family chose to commemorate, in order to demonstrate their right to land and leading positions.The relation between the small gravefield, although the biggest from this period so far found in the area of Lund, and the dominating farm, resembles that of the dominating farm at Vorbasse and its burial ground (Herschend 2009: 73-78).
Normally graves from the Lund region from this period are poorly equipped (cf.Björk 2005).The rather rich Late Roman burials from Odarslöv 51 might be markers of a new social status (founder graves) when the farm was established.After a while, their social position was obvious, and power and status was no longer manifested in rich burials.This might be interpreted as a sign of the strong control from the central place in the area, and an established and stable society (cf.Näsman 2008: 33).

The oven area
Ten ovens and four pits related to the ovens were documented some 20m to the northwest of the farm buildings (Fig. 9).These were low temperature ovens, which normally means a heating below 500 °C, which indicate food preparation (Stilborg 2002: 144).To these features should house 104 be added with ovens and pits inside it, located close to the most significant concentrations of ovens.Just to the south of this area, storehouse 114 was located, and should probably also be seen in connection with activities related to the ovens.This area was most intensively used contemporary with the farms.Some radiocarbon dates show that the area was still in use after the mid-7th century, i.e. when farm 4 seemingly was abandoned.This indicates that there still were settlements in the vicinity, perhaps to the south of the Odarslöv road.
House 104 could be interpreted as a building that has to do with food preparation, which is also evidenced by the existence of large quantities of burnt grains, mostly hulled barley, in different contexts.Most certainly, it must be understood as a kitchen, bakery and perhaps brewery.

The place and the farm
The period that precedes the Odarslöv 51 is almost devoid of finds and structures in its vicinity.We can therefore be sure that the farm was established here around AD 250.A pollen diagram from a well in the vicinity shows that it was established in an open landscape dominated by grassland and fields where barley and wheat were cultivated (Broström, Ranheden & Larsson in manuscript).It is more problematic to know whether the farm ceased after AD 615/650, or if it was relocated to a place in the neighbourhood.A clear tendency is that the main buildings as well as the halls moved from the northeast towards the southwest.This could mean that the farm was relocated to a place beneath the present-day Odarslöv road after AD 650, or further to the south.The topographic conditions for this are good in the form of a sandy plateau.There is also a neighbouring farm that was inhabited during the Viking Age some 500 m to the north (see below).The historical village site, on the other hand, is situated 1 km northeast of Odarslöv 51.
The way that the farm was organized during phases 1-4 has been briefly described above (Fig. 10).It is clear that the main building and the hall were separated during phase 1, and that the hall became integrated with the farm owner's residence during phases 2 and 3. Regarding the main building, it is clear that it was multifunctional during phase 1, but that the halls, houses 109 and 108, might have become residences of the farmer's family during phases 2 and 3.This may indicate a social marker, since it is likely that people who worked on the farm lived in a part of the main building.The sizes of the main buildings also suggest surplus production with a large byre and a great need for storage.
Osteological material that can be linked to the farm is scarce, but looking at ESS as a whole, some tendencies become clear (Magnell manuscript).During the Roman Iron Age, there seems to have been a major focus on cattle.The bone material indicates that there was a specialization in meat production rather than milk.This may indicate a specialized farm and some meat may have been used as commodities.Analysis from Uppåkra shows that animals for slaughter were imported from a larger hinterland (Magnell, Boethius & Thilderqvist 2013: 121-122).Even though the osteological material from the Late Iron Age is limited at ESS, it seems as that the specialization in cattle breeding decreased.The material points more towards a normal subsistence farm economy with a focus on extensive livestock farming.There might have been a greater concentration on milk production.The high ratio of bones from young horses during the Roman Iron Age suggests that horse breeding was also a vital part of the economy (Magnell manuscript).This corresponds well with the map from 1790 (LMV L271-6:1).According to the map, the farm was located on arable land, but facing towards meadows and wetlands.
The macrofossil material from ESS contained approximately 9000 charred macrofossils of which more than 8000 were grains from the farm at Odarslöv 51 (Broström in manuscript).The degree of conservation was good and most of the grains were identified to genus or species.Hulled barley and bread wheat represented the largest shares, while oat and emmer were present in smaller quantities.All species are commonly found in archaeological remains dating back to the Iron Age in southern Scandinavia (Larsson, M. 2013;Grabowski 2014).A large amount of the grains comes from the oven area north of the farm, and from contemporary structures.It has been suggested that this area with a concentration of ovens should be seen as evidence of specialization.The oven area was obviously used for the daily preparation of food and beverages on the farm, but it could also have produced commodities.The ovens might have been used for baking, drying, roasting, smoking and malting.Remains of ovens and the largest concentration of grain come from a pit system some 50m northeast of the oven area.This system can be dated to the Late Vendel Period and the Viking Age.This indicates continuity in the use of this particular area for preparation of food and beverage.
Close to the farm a relatively large amount of iron slag was found, most of it in connection with metal detector scanning.Slag was also found in a pit close to house 102.Most of the slag can be associated with forging, and most certainly, a smith produced the iron needed locally.As mentioned above, a burial ground can be linked to the farm, and the farm was also situated close to an old communication route.These variables are important for the interpretation of Odarslöv 51 and will be dealt with below.
The neighbouring farm in the hamlet About 500m to the northeast of Odarslöv 51 was another settlement, Odarslöv 49 (Fig. 11).The site displayed two houses representing short-lived farms during the Bronze Age and Pre-Roman Iron Age.At the end of the 2nd century AD, a small farm was established again at the eastern part of the site.In the next phase, at the beginning of the 3rd century AD, a new slightly larger house was built a couple ofm north of the first house.The houses on this farm are nevertheless small or mid-size through all phases.During this phase, another farm with a 44-m-long multifunctional longhouse was established at the western side of the settlement.However, this farm was short-lived, vanishing after just one phase.The first small farm on the eastern part lived on through a couple of phases with new buildings until it was abandoned either in the late Migration Period or in the early Vendel Period.A new small farm with two houses was established again in the southeastern part of the site during the Late Vendel Period.The farm lived on into the Viking Age (Fig. 12).A splinter of a glass vessel found in a pit shows the possibility for a small farm to access exclusive imported items.The glass might be dated to the Roman Iron Age or the Vendel Period.

The Iron Age settlement pattern in the area
To get a general idea of settlement development in the area, a study of the settlements around the central place of Uppåkra might be used (Aspeborg in prep.).The results must be dealt with in a source-critical way since the excavations around Lund are few (Jacobsson 2000;Björhem & Magnusson Staff 2006;Carlie & Artursson 2005;Carlie 2005A settlement expansion is seen in the Early Roman Iron Age.At the same time, many of the farms from earlier periods show site continuity but the houses and farms become larger.The houses often show signs of being rebuilt, which might mean that the farms had become more stationary.There is also wide variation in the sizes of houses and farms during this period, which could be interpreted in terms of an increasing inequality in society.The large farms become fewer in the Late Roman Iron Age and some hamlets disappear, especially those close to the central place, such as Uppåkra 36 (Becker 2012).This trend continued during the Migration Period and no large farms on the hamlets close to the central place from the Early Roman Iron Age seem to have survived.For example, the hall farm at Uppåkra 39, and soon the whole hamlet, ceased to exist at this point (Aspeborg et al. 2013).There are examples of both very small newly established one-phase farms and very large farms in the area, but overall, the settlements become fewer (Hellerström 2007;Runcis 1998).Some mid-size hamlets or farmsteads with roots back to the Early Iron Age a little bit further away from the central place, such as Stångby 64 and Brunnshög 12 survive up to the Vendel Period, and even into the Viking Age (Artursson 2000;Carlie 2012; Hulting Lindgren 2012a; Hulting Lindgren 2012b).One farm during the transition between the Vendel Period and the Early Viking Age might be termed large.The farm at Nevishög 38 had an impressive 37-m-long main multipurpose longhouse, complemented with a pit-house and post-built outhouses in its last phase (Hulting Lindgren 2012b).There are a few examples of farms that ceased to exist during the Vendel Period but just one, at Brågarp, that was established (Pettersson 1996).However, metal detecting around the present sites of the villages in the area around Uppåkra have yielded many Late Iron Age finds, suggesting that the settlements moved to the places of the present village sites during this time (Larsson 2003;Larsson, L. 2013).These can be exemplified by Hjärup in Uppåkra parish, Trolleberg, Svarte-Hjärup and Flackarp in Flackarp parish, Görslöv in Görslöv parish, Särslöv in Särslöv parish, Vesum in Knästorp parish, Djurslöv in Tottarp parish and Önsvala in Nevishög parish (Helgesson, unpublished).A magnate farm from the Late Viking Age, found in the village of Hjärup, also seems to underline this move to the present village site.The farm was made up of seven houses, among them the 42-m-long main building.The find material was rich and included prestige objects.A runic stone stands close to the farm.It is of "after Jelling" type and mention the man Toke who died out west (Schmidt-Sabo 2011).

Farm and village
Iron Age research in Scania, and in large parts of northern Europe, often takes its starting point in settlement remains.Time, space, organization and change are analysed on the basis of postholes, pits, wells, houses, farms, workshop areas, etc. (e.g.Berglund (ed.) 1991; Björhem & Säfvestad 1993;Björhem & Magnusson Staaf 2006).At an early stage, Vorbasse became a prototype for Iron Age settlements and villages in southern Scandinavia (Jensen 2004: 156-161).The nucleated villages with fenced farmyards and the farms lying close together were easy to grasp, since they looked very much like the regulated villages known from maps from the 17 th to 19 th centuries.The internal development of these Jutish, or western Danish villages built on traditions of rights, responsibilities, ownership of land and inheritance (Holst 2010).Although Mads Kähler Holst argues that the Jutish villages were more or less self-governed with no or hardly any external control, there is often a connection between these regulated villages and the elite farms (Holst 2010: 176;cf. Herschend 2009: 267ff).Settlement structure in eastern Denmark and southern Sweden was organized in a different way, with single farms or villages with dispersed farms (Carlie & Artursson 2005: 235, 498;Fallgren 2006).
Here the links between the villages and the prominent elite farms are missing (Holst 2010: 175f).The differences between the villages of western Denmark and eastern Denmark/southern Sweden during the Iron Age must have been that farmers and aristocracy represented different organizational principles (Holst 2010: 175-176).Scania most certainly belonged to this eastern Danish/ southern Swedish sphere (cf.Fallgren 2006;Friman 2008).This means that the villages were less regulated with a lower degree of control, and there was more room for individual initiative, and thus developments.The larger farms in most villages in Scania can probably be attributed to village headmen or other members of a less prominent elite.It is hazardous to delimit which geographical areas can be discussed when speaking about the social context of Odarslöv 51.The geographic and topographic delimitations are only physical aspects of how an Iron Age settlement and society might have been structured.Personal relationships, agreements and alliances were also necessary when maintaining, evaluating and reproducing society.Since the latter are harder to grasp from archaeological material, the former might be a fruitful starting point.A first view is the nature of the farm and its relation to the nearest neighbours.A basic question is whether the farms at Odarslöv 51 should be apprehended as a single farm or as a part of a village.
The Iron Age villages of Öland might be a good starting point to gain insight into how the landscape might have been organized around ESS.The systems with stone house foundations and enclosures are well known from Öland, and can be dated from the Late Roman Iron Age until the Early Vendel Period (Fallgren 2006: 27), i.e. contemporary with ESS.Fallgren shows that the farms usually lie 50-200 m apart in the villages of Öland (Fallgren, 2006: 77).The villages can be large and it is not unusual to have a distance of 500 m between the two farms that are furthest apart (cf.Fallgren 2006: Figures 19-22).On Öland it is possible to delimit the villages since the limestone plain, the alvar, and the marshes form natural boundaries.Graves are also often located in these borderlands.Usually it is more than 500 m between the villages (Fallgren 2006: Figures 37, 38 and 40), and this is partly regulated, as mentioned, by topography.The distance between Odarslöv 51 and contemporary houses from the smaller northern farm part of ESS at Odarslöv 49 (Fig. 11) is about 500 m.According to the situation on Öland, these farms might belong to the same village, but we cannot exclude that there are in fact two villages.One of the characteristics of the Ölandic village is of special interest concerning the large ESS farm, that is: one farm in the village/hamlet is always decidedly larger, with more buildings and larger dwelling houses, than other farms in the same village (Fallgren 2006: 77).Another find from the vicinity may qualify this picture.About 100 m northeast of Odarslöv 51 and about 60 m to the south of the burial ground, an atypical crossbow brooch was found in connection with metal detector surveys.Most likely, the brooch can be dated to the Migration Period (cf.Hårdh 2003: 51-53).It was found close to the Odarslöv road, in an area heading south towards a sandy plateau where topography and high phosphate values (Arrhenius 1934) indicate settlements.This might be the site of another contemporary farm.These facts speak for the plausibility of perceiving the farms from Odarslöv 51 as a part of a village.
Another dispersed unregulated hamlet with roots from the transition between the Late Pre-Roman Iron Age and Early Roman Iron Age has been excavated at Lackalänga, about ten km northwest of ESS.The distance between the farms furthest away is about 500 m.The farms disappear during the Migration Period (Carlie 2002a: 519ff).
The Ölandic model can also be observed in terms of how the landscape was organized.Fallgren's studies show how the fenced infields (inmark) are assembled next to the house foundations (Fallgren 2006: Figure 15) how outlying lands, in the form of wetlands and alvar, separate the villages.Seen in this perspective, cultivated land might have been adjacent to Odarslöv 51, while for example Puggängarna to the south of the farms was land.A similar picture can be seen on a map from 1790 (LMV L271-6:1) (Fig. 13).The structuring of the landscape in the mid-Roman period has been suggested in a model by Björhem and Skoglund based on the large excavations in southern Malmö (2009: 77).The farms are concentrated in village-like structures and surrounded by fields closest to the farms with meadows further away.The fields that required the most intensive management were also closest to the settlements.There is a clear zonation of the landscape and settlement units are often separated by wetlands.It is clear that the ordering of settlement and landscape was largely dictated by natural conditions, with farms, fields closest to the farm, meadows and outlying land furthest away.This can be seen in several other models and examples (cf.Pedersen & Widgren 1998: 267-333;Fabech & Ringtved 2009).The settlement structure of the Mälaren valley during the Roman Iron Age is somewhat similar to the situation in southwest Scania during the same period (Göthberg 2007).Owing to a more varied landscape in the Mälaren valley with small plains interrupted by forests, mountains and wetlands, the settlement areas are easier to delimit than in southwest Scania.Concerning the interpretation of settlement structure in the Mälaren valley, both the settlement remains and the gravefields have been used (cf.Göthberg 2000: 101-108).Whether the often scattered settlements were organized as single farms or villages/hamlets, the interpretations in favour of the later are drawn in analogy with the Öland examples and Fallgren's argumentation (cf. Fallgren 1993;Aspeborg 1998;Fallgren 2006).The interpretation of larger Iron Age settlements as village/hamlets is today more or less uncontroversial (cf. Gräslund 2004: 14).
In short, the settlement structure during the Pre-Roman Iron Age seems to be that of rotating farms.At the end of this period there seems to have been an expansion of settlements that led to the first villages.The expansion of settlements and villages continued during the Early Roman Iron Age when villages/hamlets also became nucleated and regulated since the farms tended to be more stable.The villages contained farms of different sizes, and the dominant farm was often located at an elevated position in the landscape.At the end of the Roman Iron Age and in the 4-5 th centuries, many villages disappeared, but not all.The Late Iron Age villages are often suggested to have moved to the plots of the historical villages.
Research on Iron Age settlements often highlights regionality (Näsman 1998;Helgesson 2002;Söderberg 2005;Björhem & Magnusson Staaf 2006;Fallgren 2006;Carlie 2008;Friman 2008;Björhem & Skoglund 2009;Carlie & Lagergren 2012).The starting points in these analyses can vary.Sometimes we use a retrospective analysis, where older geographical divisions are the basis, for instance the small lands, the hundreds or the parishes.Topographic variables may also be used for these purposes, and agricultural districts then often become regions.The borders or borderlands are often pointed out as mountains, ridges, forests, seas, rivers or marshes.The central places may also be a starting point.A most striking example in the vicinity of Odarslöv 51 is Uppåkra, often viewed as the dominant place the central agricultural plains in southwestern Scania during the Iron Age and centre of a political unit.
The ESS area is thus part of a community that might be defined as the plains around Lund and Malmö, and perhaps parts of western Scania.This corresponds rather well with the subdivisions of the landscape during the Migration Period that have been suggested (cf. Fabech 1993: 225;Näsman 1998: 5;Helgesson 2002: 157).

Farmer, magnate, chieftain or king
The Germanic societies in Scandinavia were hierarchical.We can grasp a development where the Iron Age societies developed from a large number of local chiefdoms during the Early Iron Age, through more complex constellations in the mid-Iron Age (tribal confederations), to kingdoms (Wenskus 1961;Näsman 1998;Helgesson 2002).This can be seen in several ways in the archaeological material.In the early stages, this is expressed through rich burials and large farms (cf.Näsman 1998;Artursson 2008;Carlie 2008).During the Iron Age, different types of central places, many of them specialized, became increasingly common.Most certainly new functions and social groups required this when society changed (Näsman 1998;Helgesson 2002;Söderberg 2005;Sanmark & Semple 2008).
At the village/hamlet level, Odarslöv 51 is clearly the dominant farm, in terms of size, numbers of buildings and its hall building.The farm owner's position as a village leader thus cannot be questioned (cf.Ethelberg 2003: 274-278).
Widgren has set up a hypothetical model for a settlement hierarchy during the mid Iron Age that is mainly based on how many animals could be stalled in the farm's byres (1998: 290).In his model, there are four different ranks, from the lowest to the highest, small size farm, medium-size farm, magnate farm and central place.Although stalling is not verified at Odarslöv, except perhaps in house 106, the sheer size of the main buildings together with the hall and other buildings definitely put the farm in his second highest rank, that of the magnate farm.
In an analysis of large farms (AD 150-550), or magnate farms, around Öresund, the size of the house plays a significant role (Carlie 2008).The total floor area of the main buildings and auxiliary buildings usually clusters between 200 and 400 square metres.On Öland, on the other hand, the magnate farms from the Late Roman Iron Age to the Vendel Period have a total floor area that varies between 558 and 834 square metres (Fallgren 2008: 70).The floor area considerably varies between the four phases at ESS, between 190 and 490 square metres.In this discussion, we must remember that hall buildings are connected with three of the phases, and that these had official, residential and ritual functions.The burial ground must be connected with this, with several well-equipped graves.The Migration Period farm, phase 3, with a floor area of 490 square metres, stands out as the largest from the period in the Öresund area.It is unknown whether this situation continued into the Early Vendel Period, but it is likely, considering the hypothetical of a fourth hall building in phase 4. The farm at Odarslöv 51 must have had a relatively high status and it was probably the seat of a local magnate.They lived off agriculture, but the possibility to produce a surplus, connected with other services, might have given the farm a special status.The long duration of the hall farm indicates that the status and wealth was inherited over many generations, which leads us to believe that this was the seat of a local aristocracy.Was the owner of the farm also a follower and ally of the leader at Uppåkra?In this context, it is interesting to notice that the South Scandinavian place names ending in -lev/ lef/löv/löf often have male names as their first element (Sofi).The ending means heritage, or hereditary estate.This means that the place-names refer to a person who left the estate to the next generation.These names are dated to pre-Viking times, i.e. the Migration/Vendel Period or even earlier (Pamp 1974: 27-28;Näsman 2012: 10; Københavns Universitet, Afdeling for Navneforskning).Hedeager attributes the -lev names to the Late Roman Iron Age and discusses them in connection with the reorganization of villages and new ideas about disposition of land and inheritance that took place (Hedeager 1993: 186-187).Tentatively a man named "Ottar" was the founder of the first farm (phase 1) and lived there (cf.Anglert 2012: 96).
The large farm at Odarslöv 51 was situated in an area that previously was considered to have been marginal.We now must revise our pre-understanding, and accept that large and significant settlements and special functions also can appear in areas traditionally apprehended as marginal.
Even if Odarslöv 51 had a relatively high status, we doubt that the place was autonomous.Investigations at Uppåkra have shown that it was the dominating central place near ESS.If we accept that a magnate owned and controlled Odarslöv 51, a chieftain or petty king most certainly ruled Uppåkra and large areas of the hinterland.The magnate at Odarslöv 51 was probably an ally and follower of the ruler at the central place.This situation might have emerged during the Late Roman Iron Age and continued during the Migration Period, when the Germanic societies were transformed into larger units.
From the region further away from ESS there are several other places to which special status has been tentatively assigned, as magnate farms during this period, owing to the sheer size of the main building (Carlie 2008).Carlie exemplifies with the magnate farms from Påarp, Annelöv, Tågerup and Dösjebro, all situated northwest of ESS and closer to the coast.However, none of these farms shows other signs of high status, such as spectacular artefacts, halls or workshops.In this context, other farms with large long houses might be mentioned, such as Uppåkra 29 and house 10 from area 13 at Mellanbyn, south of Malmö.It is possible that these farms had the same social position as Odarslöv 51.In the Lockarp-Fosie area outside Malmö, there are examples of manors from the Late Iron Age (Björhem & Skoglund 2009: 54).From the Early Vendel Period, a farm with a 62-m-long main building at Fosie has been found, but the extent of the house has been questioned (Björhem & Säfvestad 1993: 59-60).These farms have been suggested to have formed manors for a local aristocracy on a lower level than in the social hierarchy (Heimer et al. 2006;Björhem & Magnusson-Staaf 2006: 244-245;Friman 2008: 199).
All these hypothetical magnate farms, with different datings from the Roman Iron Age to the Viking Age have tentatively been perceived as local centres.They varied in importance and wealth, both from each other and over time.The aristocrats at these manors acknowledged the superiority of Uppåkra and were a part of the power structure around the central place.The rise and fall of manors from the Late Roman Iron Age to the Viking Age gives us an idea of the changing hierarchical landscape and the changing political power structure around the central place.
It is not an easy task to establish a well-ordered settlement hierarchy that mirrors the political and social hierarchy of different periods of the Iron Age.As Paul Wason puts it: Why should each level in a political hierarchy be expressed in a distinguishable level in settlement (1994: 131-132)?In the area around Lund and Uppåkra, there is a tendency for settlements to have one or more farms with large houses during the Late Pre-Roman Iron Age and Early Roman Iron Age.These farms become fewer during the Late Roman Iron Age and Migration Period.As a whole, there are also fewer farms from the later periods.Settlement continuity seems to be more common on farms further away from the central place, like in Odarslöv 51.Near Uppåkra, a disruption in the settlements during the Migration Period has long been discernible from the point of view of place names, but this is now also supported by the results of settlement archaeology.The restructuring of the farms around Uppåkra probably was a result of the creation of the Uppåkra estate, with newly established dependent hamlet/villages around the central place in the Late Iron Age (Callmer 2001).The same phenomenon has been noticed around Gamla Uppsala at the end of the Migration Period when the old villages were abandoned and instead a domain was created (Zachrisson 2013: 165-166).
The elite stratum between the chieftain and the free man The relationship between the Germanic farmers, the commoners, and their leaders is still poorly understood.Iron Age society during the lifespan of the farm at Odarslöv 51 was a stratified and hierarchical society, but also a society in transformation.There was a social stratum, or probably several, in between the leader at Uppåkra and the free men of the tribe.The organization of the Germanic army by different ranks points in that direction, as does settlement hierarchy and also the testimony of the variation in valuables, weapons etc. in contemporary graves (Pauli Jensen, Jørgensen & Lund Hansen 2003, 327;Carlie 2005;cf. Hedeager 1992: 123ff).
At the bottom of society, beneath the commoners, there were also the slaves and the freedmen (Tacitus 25).In the social structure of the Germanic people according to Tacitus, there was also the king, chiefs, warriors (belonging to a king's or chief 's retinue), priests and the people.
The retinue was mentioned already by Tacitus (Chapter 13).The retinue was based on relations of dominance, for which the family structure constituted the conceptual, cognitive framework.What is meant here by dualism is that one party had a social and economic advantage over the other without formal imperative.What further distinguishes the retinue structure is its repeatability.The structure could be duplicated again and again, but still be held together by relations of dominance (Varenius 1998: 29-30).The entire society could have been organized in that way.
The relation between Uppåkra and the surrounding "normal", and also other specialized, settlements has become increasingly pronounced in research (latest, Larsson 2014).The early central places may have based its economy partly on activities performed at the central place and partly on subordinate settlements (Jørgensen 1995).How this might have been practised around Uppåkra has been discussed from different perspectives (Callmer 2001;Anglert 2003).At the central place, the paramount elite lived with specialists (craftsmen, traders) and warriors in the retinue.In the surrounding settlements lived farmers, fishermen and craftsmen whose production supported the central place and its inhabitants.This was also where the allied elite of the aristocracy or kings lived, i.e. men of different rank, on larger farms.During the Roman Iron Age, there were hall farms close to the central place, such as at Uppåkra 39, but these disappeared during the Late Roman Iron Age.It seems as if the ordering of the landscape and the extraction of goods from the surroundings to support the central place changed around the beginning of Migration Period.Uppåkra can be described as an estate with subordinate settlements during the Late Iron Age (Callmer 2001).
At some villages in the hinterland some remarkable prestige objects from the Late Iron Age have been found that suggest a presence of an elite of a higher rank at magnate farms at those places: Skabersjö, Lackalänga, Mölleberga, the Värpinge complex and Glostorp (Helgesson 2002: 152-153, 185;Lihammer 2012: 160-204).We can probably add Lilla Harrie, Örtofta 51, to those previously mentioned.The farms seem to have been consciously placed in the landscape at roughly equal distances from each other or the central place (Fig. 14).The sites might be interpreted as seats of an aristocracy allied to the small king at Uppåkra or, perhaps anachronistically termed, as residences of his vassals (Helgesson 2002: 152-53).On the other hand, the followers at manors as Odarslöv and even larger manors were the power behind the Uppåkra leader.Be that as it may, the picture of the settlement and social hierarchy in the area through the Iron Age, with all the changes, is beginning to reveal itself even though a lot of questions still remain to be answered.The results from the excavation of Odarslöv 51 is only one piece, but an important one, in understanding the ordering and remodelling of the rulership of Uppåkra and its consequences for the hinterland.

Odarslöv in a wider context
The development of the farms at Odarslöv 51 cannot be explained only in a local or regional context.The archaeological literature deals with many different external causes on the Iron Age society in Southern Scandinavia transformed.For the period of interest one may discuss the contacts with the Roman Empire, the Marcomannic Wars, the "Migrations", the decline and fall of the Western Roman Empire, the Migration Period agrarian crisis, the dust veil of 536-537, the Justinian Plague in 541, the Merovingian expansion in the early 500s and the Slavic expansion into the Baltic Sea region during the 500s.
The establishment of Odarslöv 51 in the Late Roman Iron Age can probably be associated with changes in society.An important event is the Marcomannic Wars when southern Scandinavia was tied more closely to the continent.This affected the development of the Germanic societies of northern Europe most significantly (Jensen 2003: 377-600).Older genealogically related tribal societies were gathered in increasingly complex tribal confederations.Returning warriors from wars on the continent seem to have played a part in the transformation of society (Fabech 2011: 34;Näsman 2012: 9).New elite groups were established and power struggles among warlords led to fewer and larger social units.At the same time, at the highest level the political power was unstable and short-lived.The hegemonies were rising and falling depending of the success of the king (Skre 1998: 324f;Nicklasson 2001: 146;Wickham 2005: 371).
In the region, Uppåkra is the central place.The archaeological material from Uppåkra shows that many functions were gathered here, which normally cannot be documented from the agrarian settlements in the hinterland.It is difficult to say whether people from Odarslöv 51 took an active part in the different events that developed the society on local and regional level and whether this promoted the evolution of the farm, but seems reasonable.The farm at Odarslöv 51 shows a certain degree of status during this formative period in the Late Roman Iron Age and the Migration Period, best expressed in the large farm, the hall buildings and the rich graves.They were certainly followers of the overlord at Uppåkra, indicated by the long continuity of the farm.Otherwise, they would not have survived.
The farm might represent ten generations of different leaders from the same dynasty during its lifespan.The stability of the farm and the long use of the burial ground can thus be seen as a sign of support and subordination to the leader at Uppåkra.The central place was probably the focal point of changes of society in the region.It has been given a central role in order to describe the development during the first millennium AD in Scania (Helgesson 2002).Its significance has also been expressed in a wider northern and western European context.The place is characterized by multifunctionality, complexity, supraregionality and a complex composite structure.The finds show the importance of the site for specialized production, local and long-distance trade, and the presence of warriors; a retinue who probably had the task of defending the place, and controlling the hinterland and communications.Uppåkra was also a centre for the ritual, legal and political life, as well as the seat of the paramount elite.Uppåkra had a special status for more than 1000 years, which indicates an ability to adapt to a changing world.
From a perspective with Uppåkra as an absolute centre of the chiefdom/ petty kingdom, everything else is periphery.An alternative and more probable model is that there were places with different degrees of status and specialization in the immediate area of Uppåkra, and likely with different degrees of independence and subordination.It was a hierarchic settlement system reflecting a hierarchically stratified society.The material from Odarslöv 51 indicates a relatively high status.The gold and the glass and amber beads bear testimony of it.The gold might have been imported from Roman territory (Andersson 1995: 11), the glass beads from workshops on the Rhine (Olldag 1994) and the amber beads from Poland and Bornholm (Lund Hansen et al. 1995: 217 ff;Christensen 2011: 36).Odarslöv 51 was indirectly connected with areas further away than the Lund plain.It is not likely that the status objects were directly imported to Odarslöv 51, since many graves from the region contain similar finds.It is more probable that there were intermediaries in these transactions, and it is perhaps not a wild guess that Uppåkra played this role.This suggests some kind of bond between Uppåkra and Odarslöv 51, as well as with other sites in the region.In addition to the commodities, there might have been a flow of knowledge from the central place, crucial in a changing society.On the other hand, surplus from Odarslöv 51 in the form of horses or cattle might have been given to the leader at Uppåkra as tribute and gifts.The gold at Odarslöv 51 and other valuables might at the same time have been given to the Odarslöv aristocrat by the leader of Uppåkra since gift giving and generosity to his men and allies was expected from a good ruler.
One must also consider the strategic location of Odarslöv 51, in both a military and an economic sense, situated along the road between Uppåkra and the areas to the north and northeast.From the farm, one could supervise movements on the road from the north.In an economic sense, it lay close to resources in the marginal areas that were sparse, but necessary, at settlements on the plain, such as wood for building material and firewood.However, the economic links between the centre and the periphery, or the centre and the hinterland, in Iron Age Scania are still poorly understood.
The surplus from the farm also enabled the construction of the hall buildings.At Odarslöv 51, the three hall buildings should be seen in a ritual, social and political context.These buildings must be seen as a vital part of the ritual life on the farm but they also had public functions as the arenas for feasts and meetings of a social kind.The two later halls are much larger.It has been suggested that they mark a stage where the magnate's family moved into a residence of their own.This means that they distance themselves from the rest of the people, marking their social position.
The overall picture is that Odarslöv 51 must be considered to be a magnate farm.In relation to the smaller farm at ESS, Odarslöv 49, it furthermore must be proposed to be a dominant farm at the local level.Based on this, the magnates at Odarslöv 51 must have had a high status and rank in society.Being so close to Uppåkra suggests that it was a residence of magnates belonging to, at least, a lesser elite.An idea of the hierarchy of a chiefdom or petty kingdom can be found in the poem of Beowulf (cf. Herschend 1994: 186).At the top is the king, the foremost of hall owners, second are the other hall owners, third are the king's men, fourth and last come the rest of the men.This underlines the high status of both the farm and its owner in the petty kingdom ruled from the central place, even though there seems to have been a significant difference in status and wealth among the group of hall owners (Herschend 2001: 59-60).We may suppose that dominance and power were produced and secured by means of personal contacts.Bearing in mind the close distance to the central place the farm owner must have been a close ally of the leader at Uppåkra, a man belonging to the elite or the aristocracy.This may resemble the picture of the social structure that Ljungkvist sees around Gamla Uppsala during the Late Iron Age with a spectrum of rank between smaller landholdings belonging to a lesser elite/low ranking aristocracy to the top elite at central places like Gamla Uppsala (2006: 184, 191).In a scientific respect, Uppåkra resembles Gamla Uppsala in that they both are considered so important that the hinterland is almost forgotten and its places and people are treated as passive clients to the constantly active agents at the central place (cf. Ljungkvist 2006: 184).
In the case of Uppåkra, the people at Odarslöv 51 enter the stage as agents in history.They were agents who played an active part in shaping both the landscape and society.As magnates, the farm owners must have played an important role in the local community during a time when society underwent change and transition.That the farm could prosper for 400 years is a testimony of the farm-owning family's economic, social and political skills.The rise and fall of the farm gives us a glimpse of the dynamics of that process in the Uppåkra area.