Expansion and Contraction – An Iron Age Settlement History

3 February, 2014

This week On the Reading Rest I have no less than three excavation reports, two larger ones each with their summary in English (0):

ILLU 00Fagerlund, Dan. 2013. Äldre järnåldersbebyggelse vid Söderhällby. Inför byggnation av bussdepå. [An Early Iron Age Settlement at Söderhällby. In view of building a bus depot]. Upplandsmuseets rapporter 2013:03. Uppsala. http://www.upplandsmuseet.se/PageFiles/3820/Rapp%20S%c3%b6derh%c3%a4llby%20komprimerad.pdf

Lucas, Malin and Lucas, Robin. 2013. Gårdar och hästoffer. Järnålder och tidig medeltid I Fyrislund. [Farms and Horse Scarifice. Iron Age and Early Middle Ages in Fyrislund]. Upplandsmuseets rapporter 2013:02. Uppsala. http://www.upplandsmuseet.se/PageFiles/3820/Hellby1217_l%c3%a5guppl%c3%b6st.pdf

and a small one with no summary in English:

Hennius, Andreas. 2013. Fyrislund 6:14 Fornlämning nr 388 Vaksala socken Uppland. Upplandsmuseets rapporter 2013:01. Uppsala. http://www.upplandsmuseet.se/PageFiles/3820/Rapport%202013_01.pdf

This latter excavation concerned three large hearths and a four postholes c. 300 m south of the Hellby excavations. Fyrislund 6:14 was a small site, c. 200 m2, in the outfields or commons where activities, sometimes including fire and a hearth, evidently took place. These features may have been used in connection with husbandry (Hennius 2013:16), but strictly speaking we do not know what activities they represent, just that whatever happened took place in two periods: in the 4th-5th century at least and c. 300 years later in 8th century at least once. The site is interesting because it suggests that there were small places in the landscape that were remembered as specific for centuries. They are also interesting because the activities that they indicate were open-air outfield occupations suggesting an extensive land use and a certain pressure on resources.

The new eastern part of Uppsala, called Fyrislund, is growing up because it has become beneficial to exploit the arable land between the old and the new E4 highway, for urban purposes. The three excavations reported here, among many other ones, are situated in this new part of town. The E4 traffic corridor runs east of cnetral Uppsala serving North-South as well as East-West communications. In addition the E4 has become a ring road for Uppsala. There are in other words many profitable reasons for investing in the area. The resulting loss of an old human landscape giving way to a new one filled with outlets, fire stations, factories, service stations, bus depots, business parks, sports arenas and the odd hotel, demands documentation before the old one is erased. Despite the economic crisis, exploitation does happen and little by little a picture of the development of this 3 km2 area, which became dry land only c. 3.000 years ago, is building up. Understanding the development of the settlement is based on modern archaeological excavations. Because of all this, the area has already been commented upon On the Reading Rest (OtRR 10 December, 2012 and 18 March, 2013).

Indirectly the new excavations in Fyrislund fill-in part of the prehistoric settlement picture and they highlight the methodological peculiarities of contract archaeology. The archaeological excavation costs are part of the investment costs in the area, which became an area because of public investments in roads and archaeology. Archaeologically speaking a road further east through the woods would have been cheaper, but it would no doubt have enclosed too much land and made exploitation a visionary hazard doubling the size of Uppsala in an economic area where the willingness to invest was limited. The area that was actually enclosed was suitable and it attracted investors that may well think that archaeology could jeopardize investments.


Lucas & Lucas 2013 report two excavations where this was felt. As procedures with reference to the anatomy of the human landscape where adjustment is always in focus the trench at Övergnista is a ‘circumcision’, at Hellby it is a ‘cutterage’. Both aim to compromise. Together with the excavations at Fyrislund 6:14, the removal of a site more difficult than interesting to interpret, Övergnista and Hellby are excavations on the edge. Primarily, they represent expansion in the villages and their hinterland.

ILLU 02 & 03

The excavations at Söderhällby are brought into the discussion because Söderhällby is a ‘pure’ Early Iron Age villages and the one closest to Hellby. The chronological distribution of the 14C-dates is most similar to the Säby area in general.


This doesn’t come as a surprise given that the Säby area and the Hellby/Söderhällby areas abut, that the expansion in the whole area is late, starting in the beginning of the Common Era, and that all the sites were abandoned in the middle of the first millennium. The plateau just before 200 CE and the decline in the 4th century in Säby as well as Söderhällby, nevertheless meet the eye as rather specific similarities.

This means that we have three different kinds of excavations: two touching upon expansion areas, i.e., the margins of villages that survive the EIA, one that doesn’t, and in addition Fyrislund 14:6, which could be expected to represent expansion or pressure on the human landscape. This means that the discussion concerns settlement expansion and contraction.

As always in modern excavations dating rests with the 14C dates and among them the excavations have added 58 dates to stock of dates.

The dates from Säby or Söderhällby constitute a pattern that we would expect also from other sites in the area, but this is a pattern we don’t find when the sites are known as historical villages – Hellby since 1376 CE. (Över)Gnista since (1540) and 1334 CE. (Lucas & Lucas 2013:7-8).


Even when a site is too peripheral to have been settled as in Fyrislund 6:14), the dates differ from the normal pattern.

Although both Övergnista and Hellby were used in the Roman Iron Age (RIA) thus linking-in with the expected primary period of expansion, the decline in the 4th c. doesn’t exist and there is a small peak c. 600 CE. There were a few 7th c. dates in the Säby area too and the by chance the Söderhällby pattern may well overemphasize the small peak c. 600 CE, but it may also be a characteristic fact. Most importantly, however, are the two peaks in the Carolingian Iron Age, in the 9th c. CE and c. 1100 CE.

In these diagrams a peak means that the villagers were compelled to use peripheral village areas. There might be all kinds of reason for this, but generally speaking expansion — the need for more farmhouses – must be seen as the most acceptable. What we see in the diagrams is the pulsation of the village settlement— expansion and contraction – during a 1000 year period. Moreover, having surveyed and excavated large areas in Fyrislund – ancient monuments as well as empty areas – we may draw the conclusion that expansion in the EIA resulted in many new often short-lived farm sites. In the LIA this was not the case. Instead there was expansion in existing settlements or the successful establishment of new villages that lived-on into historical times

Indirect therefore the diagram can be describe the following way: Since there is presence on the sites there is probably also farms in the central part of the settlement sometimes during the RIA although occupation is nearly always preceded by presence. Since peaks indicate expansion and settlement pressure, lack of dates indicates the opposite. This suggests that the odd expansion in the 6th century may be a result of people moving-in from some of all the abandoned settlements. Contraction nevertheless is the hall mark of the 6th to 8th c., i.e. the PCIA, before the settlement pressure rebounds in the CIA and becomes visible in the village periphery in the 9th century. Contraction sets in again but c. 1100 CE the settlement rebounds once again.

The open site Fyrislund 6:14 fits the greater pattern inasmuch as it catches the general RIA expansion as well as the 9th c. expansion. The dates are few and may be insignificant, but patterns are nevertheless patterns.

In theory one might of course argue that that Övergnista and Hellby were established and reestablished at every peak, but it is more likely that although many RIA settlements were abandoned some were reduced and surviving the successive depressions of the 5th and the 6th centuries. Not least because there are other historical village tofts with house remains belonging to the period Roman to Carolingian Iron Age (1). Needles to say we should expect new villages to be established during the CIA peaks

If we look at settlements in terms of expansion and contraction, it is a sign of the depth of the mid-millennium depressions that not until the 9th century do the villages expand beyond their EIA boundaries.

Although we may find this interpretation convincing one could, nevertheless, come up with other interpretations and as long as none of the historical village tofts have been completely excavated, doubts as to the interpretation of the settlement development in the mid-millennium will continue to puzzle us. Contract archaeology as a methodology will not takes us into the village tofts because it is too expensive – not even if it would make it much easier to understand why EIA farms hamlets and villages look the way they do and thus lower the cost for their excavation and making it possible to prioritize in a more reasonable way.

If society finds it interesting to invest in roads and road-building techniques to facilitate development and profitable investments, then it is difficult to understand why society doesn’t invest in problem-solving excavations that would also contribute to making it more profitable to develop society. Historical village tofts will still be expensive to excavate, but EIA sites – and they are very many –  would become less expensive to excavate because knowledge about historical villages tofts would solve problems that are presently slowing down the retrieval of archaeological knowledge about settlements that were given up in the middle of the first millennium CE.


(0) A summary of the excavations at Söderhällby. In August and September of 2010, and March of 2011, Upplandsmuseet conducted an archaeological investigation of an Early Iron Age settlement at Söderhällby (Vaksala 397), on the eastern outskirts of Uppsala. The investigation was one of four major excavation projects undertaken due to the construction of a bus depot. Another settlement (Vaksala 386), located adjacent to the site of a historic village (Vaksala 383) was investigated within Söderhällby. Within Gnista parts of the historic village (Uppsala 678) were excavated as well. Finally, a burial ground (Vaksala 155) known colloquially as Inhåleskullen (‘The Devil’s Hill’) was investigated. The area is dominated by a flat agricultural landscape with hillocks comprising infields to the villages of Övre (‘Upper’) and Nedre (‘Lower’) Gnista, Över– (‘Upper’) and Neder– (‘Lower’) Kumla, as well as Norrby and Söderhällby (incl. Löt). Prominent ancient monuments include the sites of historic villages, single graves, burial grounds and prehistoric settlements. A Late Iron Age – Early Medieval burial ground, severely damaged by agriculture, in the western part of the area contains some high-status graves such as a damaged mound and a weapon grave. High-status graves were also found at Inhåleskullen, which according to the oldest cadastral maps strategically is situated in the middle of the area, on the boundaries between Gnista and Norrby. The burial ground was in use from the Bronze Age to the Viking Age. The settlement was situated on a low moraine hillock surrounded by tilled fields. Settlement remains were discovered north, east and south of the hillock. Out of a total of 630 archaeological features, 455 were postholes. The majority of these could be linked to the 17 houses or a few fences. House types included 14 larger or smaller three-aisle longhouses, four corner-post houses and one two-aisle house. About 25 postholes and hearths were discovered during construction work north of the investigated area. Other features included 92 hearths, five ovens, three cooking pits, two tar making pits and a small number of pits, dark patches and stone features which were difficult to interpret. The find material was very limited and dominated by about 1 kg of animal bones. These represented the most common domesticated species. The horse remains were interpreted as remains of sacrificial feasts. Another interesting find was a quern stone recovered from a posthole in a house situated on the hillock, probably a trace of local cult by way of a sacrificial offering. The houses have been in use on two co-existing farms, one situated in the north, the other in the south and southeast. A row of postholes on the hillock may have functioned as a boundary between the two farms. Radiocarbon samples were mainly collected from preserved wood, predominately pine, in postholes. Through dating of all the houses, a chronological development has been outlined (house numbers within brackets): Phase 1: At the onset of the Early Roman Iron Age, the settlement consisted of a single farm, represented by a multifunctional dwelling (3). Phase 2: Later on in the Early Roman Iron Age, house 3 was replaced by a very substantial house (16), situated further to the south, on the crest of the moraine hillock. At the same time, another farm was established south of the hillock, with a dwelling (4) and a large storage building (10) at right angles to each other. A row of postholes on the hillock may have formed part of a boundary between the farms. The elevated position of house 16 may indicate a certain measure of influence emanating from the north farm. Phase 3: Early on in the Late Roman Iron Age, another house (2) was added to the north farm, possibly replacing house 16. In the southeast a dwelling (7), was erected, probably along with a smaller outbuilding (9). Phase 4a: The period of most intense settlement activity, dating from the latter part of the Early Roman Iron Age. To the north, house 2 was replaced (1). Also, in the south a dwelling (13) was erected, probably along with a storage building (5). Evidence suggests that the south farm was home to two households during the Later Roman Iron Age, with houses 13 and 7 coexisting with house 1 to the north. House 13 was very similar to the earlier house 16, partially situated on the crest of the southern edge of the hillock. Phase 4b: House 7 was replaced by or converted (8), while another was erected in the southeast (6). There seems to have been a continued presence of two households in the southeast at the close of the Late Roman Iron Age, while the settlement to the north seems to have been abandoned and replaced by a two-aisle house (14) of unclear interpretation. It probably represents an outbuilding to the two farms or households to the southeast, indicating a clear reorganisation of the settlement. Phase 5: Only one farm and one household seem to have remained during the Migration Period. Its main building (12) replaced house 8 on exactly the same location. Other parts of the settlement were used for other activities, judging from the presence of a corner-post structure (15) on the site of the north farm, and a simple three-aisle house (11) in the south. Probably, the remaining farm had access to the entire settlement for various activities. The smaller houses could also have been outbuildings located on the old farm site and used by its former inhabitants. There is no dated activity whatsoever after the transition from the Migration to the Vendel Periods. The Söderhällby settlement seems simple and well-defined, based mainly on selfsufficient agrarian means of subsistence. On the other hand, the settlement is one of many in a landscape of complex ownership structure, judging from the cadastral maps. Also, material from some graves in the area indicates the presence of a local elite, possibly linked to some form of manor, or large estate. This layout is echoed in many parts of Uppland around the same time. For instance, seven excavated settlements in Säby, Danmark parish, seem to form part of a village community which may be subject to a local authority. A central manor as described above can hardly have existed on the Söderhällby settlement which was abandoned after the Migration Period. Around the same time the village known from historical sources emerged on a new site and at Övergnista. A stricter management of the landscape may be the reason for the changes and an underlying social organisation can be discerned. Later on, during the Viking Age, stricter property boundaries may have emerged, and new status graves were erected on a burial ground to the west, on land belonging to Kumla, Söderhällby and Norrby. The burial ground was used until the Medieval Period, and some graves had Christian attributes. It is possible that the divided ownership structure evidenced by the cadastral maps can be traced to a second reorganization during the Medieval Period, linked to the abandonment of the burial ground.


A summary of the excavations at Över Gnista and Hellby. From early September to mid-November 2010 Upplandsmuseet conducted a series of archaeological investigations in Fyrislund, a rapidly growing industrial estate on the eastern outskirts of Uppsala. The investigations were necessitated by the Uppsala City Council’s plans to build an access road and extend amenities to the new development. An initial evaluation of the sites in September unearthed prehistoric settlement remains adjacent to two small hillocks where the historical villages of Hellby and Övergnista had been situated. Övergnista had no recorded settlement after the mid-17th century whereas Hellby (later known as Söderhällby) was inhabited until the 1950s. In October and November major investigations were undertaken on both sites. The Hellby and Övergnista sites were designated as Raä Vaksala 386 and Raä Uppsala 678 respectively in the Swedish National Heritage Board database for archaeological sites and monuments (FMIS). At Hellby a total of 189 features were found and these included postholes, hearths, ovens and various pits. A total of ten houses were found, including seven one-aisle, two three-aisle and one simple corner-post structure, as well as parts of an enclosure system. The site’s earliest activity dated to the Roman Iron Age, when a couple of hearths and an oven seemed to have been in use. The houses were erected later in two separate phases, the first from the first centuries of the Viking Age (c. 8th – 10th cent. AD), the second from the late Viking and Early Middle Ages (c. 9th – 12th cent. AD). Small amounts of animal bones, mainly cattle, horse, sheep and pig, but also very small amounts of canine bones were found, some of which may have constituted ritual offerings deposited in postholes. Other finds included a few ceramic shards and a clay mouthpiece from a bloomery furnace, possibly also deposited in a posthole as an offering. Iron objects included a ring from a bridle and a ship rivet. These, and other more latter-day metal objects were found in the topsoil during the initial metal detector sweep of the area. At Övergnista a total of 179 features were found and these included postholes, hearths, various pits and areas with cultural layers. A total of eight houses were found, five one-aisle, one three-aisle and two corner-post structures, as well as parts of two enclosures. The site’s earliest activity dated from the Roman Iron Age, when a couple of hearths were in use. The houses were erected later in three separate phases, the first dating from the Vendel Period (c. 5th – 7th cent. AD) the second from the Viking Age (c. 7th – 10th cent. AD), and the final phases from the late Viking and Early Middle Ages (c. 11th – 12th cent. AD). Finds included bones of cattle, sheep, pig, as well as some canine bones. Other finds included ceramic shards from a complete but fragmented vessel and a fossil, possibly a sacrificial offering. Iron artefacts included hob nails, knives, chisels and ship rivets, and an amulet in the shape of a strike-a-light. Also a bronze weight was found. Most metal objects were discovered in the topsoil during the initial metal detector sweep of the area. The most spectacular find on the site was the almost intact skeleton of an adult and healthy horse, dating from the early Middle Ages, which been deposited on its back in a Late Viking Age house, presumably while this structure was still standing. The horse was interpreted as an offering, possibly a closure rite, commemorating the abandonment of the house or even the abandonment of animal sacrifice itself following the advent of Christianity. The excavations at Hellby and Övergnista increased the number of known Late Iron Age houses in the Uppsala area by 20-30 percent, placing them among of the largest known settlements from the period in the area. Compared to other historic villages in the area, Hellby and Övergnista were situated on relatively low elevations, and not adjacent to any prehistoric grave fields, which can be viewed as an indication that they were somewhat later settlements, established on old grazing land, only recently made dry and habitable by shore displacement. Another unusual feature of these villages as well as a few others in the area is that they had two separate village tofts, as evidenced by cadastral maps and medieval documents. Also, the cadastral maps show that there was much intermingled ownership of land among these villages, suggesting that they might once have been part of a much larger unit, possibly a manor or another form of settlement belonging to the upper ranks of Late Iron Age society.

(1) Ros et al. 2008.  Jonas Ros, Pierre Vogel & Tony Engström. Mörby: Järnåldersboplats, historisk bytomt och skålgropslokal Förundersökning och särskild arkeologisk undersökning RAÄ 246, 255, 459 och 473, Mörby 5:1, Turinge socken, Nykvarns kommun, Stockholms län, Södermanland SAU Rapport 2008:12. Uppsala. http://www.sau.se/filarkiv/rapporter/2008_12%20morby.pdf

One Response to “Expansion and Contraction – An Iron Age Settlement History”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: