översiktskartaThis week On the Reading Rest I have three volumes from the Norwegian E18 project, the contract archaeology carried out between Gulli and Långåker. The volumes are impressive and they are part of the significant development of Norwegian contract archaeology that has taken place because of the many carefully designed excavation projects in connection with the new highways southeast and southwest of Oslo (0).


Gullilangåker framsidorMjærum, Axel and Gjerpe, Lars Erik (eds). 2012. E18 prosjektet Gulli-Långåker. Dyrking, bosetninger og graver i Stokke og Sandefjord.  E18 prosjektet Gulli-Långåker, bind 1. Oslo Fagbokforlaget.

Gjerpe, Lars Erik and Mjærum, Axel (eds). 2012. E18 prosjektet Gulli-Långåker. Jordbruksbosetning og graver i Tønsberg og Stokke.  E18 prosjektet Gulli-Långåker, bind 2. Oslo Fagbokforlaget.

Gjerpe, Lars Erik (ed.). 2012. E18 prosjektet Gulli-Långåker. Oppsummering og arkeometriska analyser. E18 prosjektet Gulli-Långåker, bind 3. Oslo Fagbokforlaget.

There are almost 1,000 pages to read, but I have chosen to look at the Hørdalen Area, Site 51. This site is a field system containing cleared fields, intensive soil improvement, a droveway alignments, clearance cairns, linear cairns, etc, etc.

This means that I read two chapters:

Mjærum, Axel. Dyrkningsspor og fegate fra eldre jernalder på Hørdalen (lok. 51) In: E18 prosjektet Gulli-Långåker. Jordbruksbosetning og graver i Tønsberg og Stokke.  E18 prosjektet Gulli-Långåker, bind 2. Acronym: MjAx

Cannell, Rebecca J. S. The application of multi-elementalanalysis at Hørdalen (site 51): an evaluation of methods and results. In: E18 prosjektet Gulli-Långåker. Oppsummering og arkeometriska analyser. E18 prosjektet Gulli-Långåker, bind 3. Acronym: CaRe

And the relevant section, pp. 132-147 in:

Svensson, Nils-Olof and Regnéll, Joachim. Vegetationsdynamik och merkanvändningshistoria längs vägsträckan Gulli-Langåker i Vestfold. In: E18 prosjektet Gulli-Långåker. Oppsummering og arkeometriska analyser. E18 prosjektet Gulli-Långåker, bind 3. Acronym: SveRe

The early surveys and investigations at Hørdalen, followed-up by the project designed by the contract archaeologists and the archaeometrists involved, have resulted in stimulating results that add to our knowledge and open up a window towards a more hypothetical discussions, which may be interesting in the future when it may become possible to investigate field systems and farms united as one context. In the following I will look at the results from Site 51 in relation to such a future scenario.

site 51 surfaces

The Hørdalen area and Site 51 has been understood as a gentle W-E slope characterized by nutrient-poor soils consisting of a slightly higher A-Surface and consequently a slightly lower B-Surface. Despite the soil qualities this slope has been developed into a field system used for more than a millennium. In many ways earlier analyses have been confirmed and become less monothetic owing to the new investigations, but archaeological features, chemical multi-element analysis and paleobotanical analyses have shown that the border zone between surfaces A and B must be understood as an area in its own right dominated by clearance cairns and an oven area. In fact the AB Surface and the western part of the B Surface, which is dominated by cairns and the oven, can be seen as a transition from surface B to surface A. I will treat it as AB+B Surface as a unit creating a triparted site. Site 51 represents an agricultural space developed and used for centuries in different more or less intensive ways. The Hørdalen Area of which site 51 is just a part was probably used continuously, but when singling out site 51 a more varied usage comes to the fore.

jämför Gausel Hördalen

Before the beginning of the Common Era, the curves are quite similar at Gausel and Hørdalen as indicated by the parallel green lines inserted into the above illustration. It is conceivable to explain this similarity by suggesting that the investment in labour has led to a certain outcome – increase in investments leads to an increase in productivity. The period of investment belongs to a period in which Bronze and Early Iron Age farms are not yet stable, but nevertheless linked to a specific settlement area and thus able to use the same field system reaching it from different farm sites, some being more optimal in relation to the system than others.

In the beginning of the Common Era and during the following centuries, investments at Hørdalen drop, but at Gausel the outcome represented by the charred seeds found in the Gausel excavations, grows. Tentatively we may conclude that adding nutrients to the fields more effectively, for instance bringing livestock to the fields at intervals and collecting more dung in the byres, may explain the difference. Bringing manure into the fields from a stable nearby farm seems more manageable than bringing it from less stable farms with a variety of situations. However, bringing farm hands to a field system from different farms, investing labour in burning and slashing as well as in clearing the site, seems a logistically straightforward possibility already in the Bronze and Pre Roman Iron Age. Since transportation costs are low when clearing fields and making ashes on the spot, it is more easily organized than bringing in nutrients, which is expensive in terms of transportation.

Since Gausel represents a larger area than Hørdalen Site 51, which may represent nothing but itself, the seeds at Gausel probably had more sources other than just one field system. Hørdalen, therefore, could have been given up for all kinds of irrelevant reasons unknown to us, but if for the sake of the discussion, we suggest that it is more likely to fit into a general pattern, as indeed it did up and until the beginning of the Common Era, rather than a deviant one, then the expected change in farm size and settlement stability in the beginning of the Common Era could indicate a shift in labour investment from developing visible clearings to adding ‘invisible’ nutrients. Moreover, with stable farm sites, slash and burn farming will be more difficult to imagine given the deforestation that will take place around a farms with a stable position in the landscape.

Site 51 Hørdalen, with the droveway on the B Surface, indicating a nearby farm, probably 80m north of the excavation area (MjAx:187 with reference), has a place in the human landscape as well as a formation history. If we analyze the 14C dates the formation can be termed additive as well as varied.

surface dates hördalen investments

Given the dates on the different surfaces (and I do not take the problematic optically stimulated luminescence dates, the OLS, into account) the formation of the site is additive inasmuch as Surface B is the first to be developed, followed by extensions, Surfaces AB+B (framed by a green line on the map) and A. It seems significant that little by little and up and until the beginning of the Common Era investment periods gradually become shorter and more intensive. They are indicated by green shades in the above figure. Moreover, it was uncommon that investments peaked in two surfaces at the same time. As MjAx, CaRe and SveRe all point out, the site is intricate and time seems to be a decisive factor when it comes to the complexity of the site. It seems that when the system has been created it may have been and probably was used for many different agricultural purposes.

Although CaRe is cautious in her conclusions it stands to reason that the A Surface has received more nutrients as a result of anthropogenic activities than the B Surface. As a correlation to this surface character, the 14C-dates indicate a chronological dimension to the surface division suggesting (1) an earlier period of investments into the construction of cairns clearing of fields and (2) a later period in which manure was more important and investments thus less visible. As SveRe shows the fields were not used continuously, but alternating with periods when animals grazed and fertilized the fields.


The project designers were well-aware that they would not be allowed to link the excavation of the field system with the nearby settlement. Had this been possible the contextual knowledge produced would obviously have been more important and less expensive owing to the research design being more rational. But as long as the Scandinavian countries, where governments have copied each other’s legislation for under hundreds of years, are not prepared to fund a percentage or two of the national cost for contract archaeology to further archaeological development projects that will enhance our knowledge and make contract archaeology less expensive, archaeologists are not allowed to do better that the E18 project. Given the attitudes of heritage boards this Project level is excellent, but nonetheless a matter of little by little setting new archaeological standards whenever Chance and Circumstance makes it possible.

The reason why it is so uncommon to excavate the farm and its fields, 80 metres apart, has to do with the notion of the ancient monument. Up and until WWII ancient monuments were well-defined physical structures preserved into modern times and thus incredibly marginal inasmuch as nobody had bothered to erase them from the surface of the earth. From the 1960s and onwards the invisible human landscape and its ancient monuments, not least from the Iron Age, were recognized, and by definition an ancient monument, instead of being a monument, became an area looking like plough land, small forest and meadows. The most common monument became hectares in central parts of the modern human landscape. The past became almost as large as the present covering large areas with no definite borders. At the same time the past became unspectacular. With hardly any artefacts to show let alone grave goods the Iron Age became full of social science, subsistence economy, social structure, functional and symbolic architecture, etc. The familiar past, that is, the traditional Iron Age, could no longer be mirrored in monuments of death and thus it passed away resurrecting itself as everyday life rather than funeral feasts and burial rites. The past became loaded with its own history, kept alive for instance in field systems that had been used, developed and known for a millennium already in the beginning of the Common Era. No longer monumental, the past started to grow and its contexts became hectares rather than square metres. The heritages boards are unable to cope with the consequences.

Archaeologists know how to excavate the past, but they are seldom allowed to do it. And that is why large parts of 30 and 20 years old excavations stand out as a waste of money and most importantly as a legacy of heritage management. Many excavations have been wasted because they were never reported or because they have been buried in inaccessible archives, but quite a number are pointless because they were, from the very beginning, focussed on pointlessness.

(0) A general discussion about the multi-disciplinary design of a large contract archaeological project and Gulli-Langåker in particular can be found in:
Gjerpe, Lars Erik. 2013. Om arkeometri, en fornøyd arkeolog og jordbruk i eldre jernalder. Primitive tider 15. Pp. 33-57.


Fig 00 översiktThis week On the Reading Rest I have a report in Swedish. The language is befitting. Prehistoric segregation is an important phenomenon easily niched in the Swedish landscape. Owing to the isostatic uplift that started when the ice sheet disappeared, settlements often sort themselves according to productivity and the less successful sit at higher altitudes than the successful ones. In Swedish, therefore, the present report presents four not very impressive sites in such a segregating Swedish landscape. Today, and since the Iron Age, they are situated at the margins of a relatively prosperous rural area.

FASEALASAndersson, F. et al. 2013. Fredrik Andersson, Susanna Eklund, Ann Lindkvist & Anneli Sundkvist. Med bidrag av (with a contribution by) Anneli Ekblom, Rasbo-Hov: From Bronze Age Shores to Iron Age Fields. Särskilda arkeologiska undersökningar inför omdragningen av väg 288, Jälla – Hov Rasbo 632, 634, 635, 659, 660, 668, 704 & 705, Hov 1:7, Uppsala kommun, Uppsala län. SAU rapport 2013:4.

I abbreviate it FASEALAS [Fa-sea’-las] and have based all illustrations upon it. Most of the illustrations are adaptation.


Along the present road section the very first trace of human activity is a cooking pit situated on a small island on the site Raä 634 (Andersson in FASEALAS). This island must have been a stopover for people entering or leaving the passage that begins or ends just north of the island. The place would have made a perfect camping site, a terrace on the beach below and between two small hills. If you wanted to sit in the afternoon and evening sun cooking and enjoying a meal a summer’s day in the 18th c. BCE pondering whether the Neolithic was drawing to an end or Bronze Age commencing, this was the place to sit.

In the millennium around 1000 BCE, the larger area to which the sites in this report belong, is characterized by two kinds of easily recognizable monuments: burnt mounds and cairns. Both are built and heavily dependent on stones either un-burnt in the cairns or fire-cracked and brittle-burnt in the burnt mounds. The other elements involved are air or earth in the cairns, but in the burnt mounds in which there is never any air, the earth is mixed with all kinds of remains related to all kinds of human activities. In the cairns the remains have to do mostly with burying humans.

Materially speaking burnt mounds are broad-spectrum and built over a considerable period of time with a large element of ´produced’ material, while cairns are narrow-spectrum and built during a short primary building phase with material that was ‘gathered’. Burnt mounds are often related to a boulder, cairns less so. Both monuments have a relation with the landscape. Cairns are often easily visible at a distance and often vantage points, while burnt mounds belong to small scale environments. Both monuments share a relation with waterways, but in a bay environment suitable for landing a canoe, and situated as an interface between productive lands and sheltered waters, the burnt mounds are more frequent than the cairns. Both monuments cluster, but as Ann Lindkvist’s analysis shows, clusters of burnt mounds are not as peaked as cairn cluster.

CLUSTERThere are two reasons for this. Burnt mounds are closely related to a production landscape and their clusters will reflect the size or productivity of niches or areas in the landscape. Cairns are related to the natural landscape and their clusters will reflect points in this landscape – visible points or vantage points.

In the eyes of human beings, seeking primarily to sustain themselves, the isotactic uplift in a landscape such as that of the Northern Mälar Valley will change a given environment from a prominent point in a natural landscape to a prominent niche in a productive landscape. This change will cause clusters of burnt mounds and cairns sometimes to overlap, but most often the fact that on a dwelling-site or local scale they do not overlap is the more striking.

Burnt mounds as well as cairns are connected with remains of cremated human beings. Cairns, which are often secluded monuments, have singular grave qualities, while the cremated remains in burnt mounds demonstrated ways of including the remains of the dead in the daily life of the living. The latter context, the daily interaction between the living and the dead, involves a third phenomenon: the fact that boulders attract the remains of cremated human beings as well as the produce of daily life. It would seem that if possible a dwelling site will involve remains of cremated human beings and ritual sites. Rather than being confined to graves the remains of the dead are integrated into the monuments as well as the life of the living.

Lastly it must be pointed out that building a cairn or a burnt mound is not a primarily something one does to sustain oneself, both monuments as well as the activities around boulders, represent ritual needs in prehistoric society. Cairns represent sporadic events, while bunts mounds refer to frequent events.

Access osvTo get a general feeling for the landscape when it starts to attract people we may compare the clustering tendencies among burnt mounds and cairns respectively to the isostatic uplift. This demonstrates that the cairn clusters in the northeastern part of the larger area have to do with the access from the north and a landscape in which the shores have a sharp inclinatrion and heights suitable for cairns. There is little fertile land in this area and thus no burnt mounds. In the western part of the greater area the picture is reversed, access is from the South. The concentrations of burnt mound in the northwest are situated along to a bay behind a narrow sound. On the shores around the bay it is easy to make landfall and the soils are fertile.

In the central part of the area, which is accessed from the south, the cairn and burnt mound landscapes overlap. And the area should be understood primarily to consist of three phases of usage governed by the isostatic uplift. The northernmost part is accessed through narrow sounds which dry up early especially the eastern water way. The southernmost part has the longest history because it is a broken landscape that attracts burnt mounds as well as cairns, probably at different points in time.

Between these two landscapes there is a small intermediary one situated at the southeastern passage to the northernmost landscape. This passage dries up c. 1200 BCE creating the shorelines followed by the new road, which in its turn caused the excavations to take place. During the following centuries, shorelines rapidly disappeared southwards creating a flat and wet landscape difficult to access. Two sites with burnt mounds, Raä 634 and Raä 668, nevertheless survive as coastal into the 8th c. BCE.

Rasbo-hov 800 BCE

Inlet and cemetery

Originally Raä 668 belonged to a small cluster of burnt mounds facing north towards the bay accessed from the southeastern passage. We may in other words conclude that c. 1200 BCE the isostatic uplift brought an end to a small marginal niche in the maritime Bronze Age landscape. Circa 1000 BCE Raä 668 was revived, but this time from a shoreline southwest of the monument because the bay to the north could no longer be accessed. At the same time even the burnt mound at Raä 634 was built on a shore facing southwest. Consequently, the changing landscape in tandem with the traditional subsistence system favouring coastal settlements created two marginal 9th c. sites. Both Raä 634 and 668 are Bronze Age settlements, without house constructions since the house remains found at the sites and situated exactly where one would prefer to build one’s house, are Iron Age.

Although deceased people were taken care of on sites characterized by burnt mounds, a small cemetery, Raä 659, was added to the landscape c. 1100 BCE situated in a cove a couple of hundred metres west of Raä 668 and 1.5 km east of Raä 634. This site was almost overlooked by the archaeologists responsible for survey and trail excavations and thus also by the heritage authorities at County Board, None the less this cemetery is by far the most interesting site in this part of the road section and accordingly, loosing information about it, while gaining conventional knowledge about a host of others, would have seemed a very high price to pay for conventional knowledge. Although the County Board was unable to take the right decision: the total excavation and precise chronological understanding of the whole cemetery — archaeologists eventually managed to record and excavate some of the monuments. This allowed them to conclude in general terms that the site was a peripheral burial ground used occasionally during c. 1500 years. Over time, its position changed from one next to a sound, over one in a cove next to a possible landing place (completely destroyed by road building) into becoming an ancient burial ground revived c. 500 CE. By then it was situated on a shelf below a ridge at the edge of the forest. The site is a diminutive ritual place of a surprisingly long-lasting importance to prehistoric man.

Rasbo-hov 400 BCE

During three centuries the water withdraws causing the section where the road runs today to become inland and a more and more marginal part of a flat agricultural country around a center with low hills, a country surrounded by peripheral. But owing to a change in subsistence economy, which allows smaller family groups to survive during a longer period at a given site, the three sites are reestablished in the Pre Roman Iron Age and a new site is added (Eklund in FASEALAS). Raä 635 is a farmstead next to suitable pastures and arable land. Raä 634 gets an annex, Raä 632, bordering on the southern side of the same arable land. These small settlements are not permanent. On the contrary they are occupied at intervals for no more than a house generation or two. The farm houses are small, but the sites are used by minor households from the 5th c. BCE to the 6th c. CE. The four sites Raä 634, 632, 635 and 668 represent locations where is it possible at intervals to settle a family or at Raä 635  two families (Eklund in FASEALAS). Probably these settlements have kept the cemetery Raä 659 in living memory. At least some graves on this cemetery belong to the 6th c. CE.

Given the ritual practices of the period, it is worth asking oneself, whether the remains of one of the three human being in connection with the block A1342 (In total c. 370 grams of bone) on the cemetery Raä659 (Eklund in FASEALAS), dated (Ua-40848) 2164 ± 31 BP (360–110 f.Kr.) is the same being as the remains (40 grams of bone) in connection with the block A4149 on the settlement Raä 668 (Lindkvist in FASEALAS), dated (Ua-41796): 2199 ± 37 BP (380–180 f.Kr.). The question is rhetorical, but nevertheless worth posing.


At one level the story about the sites in the road section is a function of the general change in more dynamic parts of society, at another it is a distorted reflection adding a critical light to the general picture. At a third level we get a glimpse of peripheral social values.

Although some of the places are very old in their landscapes they are neither abandoned nor developed – they are simply open or common places befitting needs that may occur in any society. Outside the road section and perhaps in the immediate vicinity of the excavation trenches, there are perhaps more populous and lively places characterized by the commonplace, but the sites as we know them were nevertheless unimportant and undecided during thousands of years. This doesn’t mean that there was no drama here – actually both House 1 and House 3 at Raä 635 were  burnt down (Eklund FASAELAS:146) – it just means that so little happened on these settlements that the life lived there didn’t bother to erase the traces of an event as dramatic and yet as commonplace as a fire. Instead of tidying up the scene and rebuildung House 1, which occupied the best location at Raä 635, people preferred to build something next to the ruined house.

Rasbo till Hov dateringar

If we sum up the dates related to these outskirts there are two intensive phases c. 800 and 300 BCE and a long less intensive phase covering the first five centuries of the first millennium CE. In addition to these phases there is occasional and sporadic presence signified by the odd 14C-test. There are no remains of buildings to counteropint the burnt mounds of the intensive presence c. 800 BCE, but c. 300 BCE the first house remains start to occur. This shift is a reflection of the common shift in subsistence economy, but socially speaking settlements are still short-lived exploitations of resources. In the beginning of the Common Era something new, a function of general social change, starts when subsistence economy in the road section gains sustainability and when some new settlement locations are added to the old ones. If this period, 0-500 CE, had been the most intensive, the 14C-diagram would have resembled that of the prolific Roman Iron Age farm, which often occupied a place that had been settled several times before it became the location of a permanent and successful farm. In the road section the typical picture is in other words distorted. Contrary to success, the persistent low-frequency presence indicates that the periphery works as an expansion vessel, albeit in a settlement rather than hot water system: when settlement pressure in central areas becomes too high and the settlements too crowded, people drop off into the outskirts and margins of rural landscape and society.

We may look upon this expansion as something positive suggesting that indirectly it shows the right to found a family still to be respected. There is little doubt on the other hand that this right was never meant to result in social segregation and since segregation is at play, we can easily imagine that not everyone living in the outskirts enjoyed a model family life – if one can live in an outhouse in the Eketorp ring fort (OtRR 25 Nov 2013) one can do so too in the settlement at Raä 635. In the long run, inability to prevent segregation from growing will harm society (O).

Raä659 There are obviously too few graves in the road section, if we expect those who lived there to be represented in monuments, but more importantly there is the uncommon cemetery, used from the Early Bronze Age, EBA, to the Pre Carolingian Iron Age, PCIA. It would seem that this cemetery was inaugurated by a stone setting representing a ship at the shore of the EBA sound or LBA cove.

Although this monument is at the bottom of a slope it dominates the cemetery in a way quite opposite to the normal monumentality, which could well have been a cairn or two at one of the obvious locations in the immediate vicininty. There is nothing wrong with the symbolism of the ship, given the BA landscape, but scunnering hights and cairns or diminishing their value is deviant BA behavior and almost impossible to excuse with reference to building a cairn being too much labour for a few outskirters. On the contrary, the boat-on-the-shore cemetery seems a most significant contribution to a new and this case outskirt identity in a shore-bound newly marginalized area (cf. Eklund162 ff.). A break with traditions, it seems significant that the site continues to be part of the human landscape even after the water disappeared c. 600 BCE (cf. Lindkvist 2013:153;fig 119). The road section Rasbo to Hov runs through the traditional lands of the prehistoric outskirters — the Commons?.

(0) Since this topic has been relatively frequent On the Reading Rest, I have added the category ‘Iron Age Segregation and Poverty’ to the existing ones.

A Fit of Formalism

14 October, 2013

Today On the Reading Rest I have a report in Swedish on excavations south of Uppsala in the Campus of the Swedish Agricultural University at Ultuna.

Ultuna fig 00Huldt, Helena. 2013. Att återvända. Arkeologi I olika skeden från Södra Gärdet I Ultuna. SAU rapport 2013: 6. Societas Archaeologica Upsaliensis. (Returning. Archaeology in different phases from Södra Gärdet in Ultuna).

This site area is interesting for two reasons. (1) Once it was one of the manors established in the nucleus of the Uppsala kingdom in the 6th century. (2) In modern times it has been the playground of the government authorities establishing and subsequently expanding the agricultural university. In this process, The Heritage Conservation Act didn’t bother the executive authorities until the County Administrative Board (CAB), and not least one of its antiquaries, came up with the novel idea of asking the agricultural university to follow the law. The Ultuna heritage experience has made excavations difficult, fragmenting the monuments, and one of the great achievements of the present report and others from later years is the competent way in which they show how archaeologists have manage to bring some order into a rural context destroyed by 160 years of campus building.

As always when archaeologists investigate an area they come across something unexpected in addition to the expected. This time in Ultuna it was a long row of hearths from the Late Bronze and Early Iron Age. (LBA & EIA) These hearths were situated in an open terrain parallel to the southern shore of a once small peninsula.

Ultuna fig 01

Seventeen of the 45 hearths are 14C-dated, but since they belong to the transition from the LB to the EIA, the plateau in the calibration curve make most of them look contemporary (1). Two dates are contaminations hundreds of years older and younger than the 15 central values.

Ultuna fig 02

Among the fifteen dates there is an early outlier date suggesting a start c. 650 BCE, and there seems to be a gap in the central probability values c. 550-490 BCE.  On average in each end of the row the dates are younger than the bulk of dates in the centre. The average before present 14C-year in the centre is bp 2446 and in the margins it is bp 2404. This difference defines the time gap in the 6th c. BCE and we may suggest that activities in each end of the row were begun and brought to an end later than the similar activities in the centre. This means that the row was formed during more that two centuries when the hearths reused. Despite their long use and period(s?) of construction and maintenance, it meets the eye that they are so regularly spaced that one may wonder whether this impression is the result of chance or intention.

Now and again in Scandinavian prehistory there are times when material expressions of formalism seem to become important. A case in point is the 4th-5th c. CE when widely different phenomena such as written poetic metre and measurements in buildings were formalized. It may be argued that formalizing architecture, standardizing the foot, was rational — in addition giving architecture an aesthetic quality — but the metre was clearly an aesthetic way of writing and talking standardizing sound and prosody in a distinctly non-Latin way. Germanic rather than Latin, defining an epic and a lyrical metre was nevertheless an adaption of Latin poetic diction – a Germanic echo of Latin verse.

Employing formalism as guidance rather than simply reproducing functional structures, is a powerful mode of thinking, closely related to mathematics, often striking humans as linked to divine order despite its human form and origin as a description of the world. That is why we tend to consider formalism to be related to progress and rationality as well as linked to the ascendency of a civilization. But we are also afraid of overdoing it because undue formalism will obstruct rationality. Although formalism may inspire arts as well as making art commonplace it stands out as a prerogative for an outstanding civilisation and for that reason we look for it also in the past. Showing formalism to be recurrent rather than progressive is part of the archaeologist’s critique of simplistic history.

Chance, function, intention or formalism – what’s behind the Ultuna row pattern of hearths?

Chance, i.e. randomness, can easily be ruled out, inasmuch as following a shore line is not random behavior. But aleatoric chance, i.e. randomness among a limited number of outcomes such as 1, 2, 3, 4, 5 or 6 when throwing a dice (L. alea) – a function of its structure – could still be imagined. However, looking closer into the hearth pattern it becomes apparent that the row is sometimes broken, and if we count the number of hearths between the breaks, it becomes evident that the sections consist of four hearths or a multiple of four. In addition, there may be a few pairs, rather than quadruples, in each end of the row. This part of the pattern points to function i.e. structure and intention.

Ultuna fig 03

There is not much in the hearths themselves that points to their function other than controlling the fires that were burning on them and it is thus impossible positively to say whether they burnt for light or cooking or heating or a combination of functions. Their construction nevertheless is relatively standardized given later disturbances by the plough. They are constructed as shallow pits lined with stores and quite often the lining end in a ring of kerbstones marking the hearth on the ground and making it manifest.

Ultuna Fig 04Many hearths show signs of having been used repeatedly as well as maintained and it would have taken many more 14C dates to understand the time depth of each hearth and of the system as such. That would have been interesting to know, but as always in contract archaeology, science and research are completely dependent on the benevolence of the CAB when autocratically it sets the economic — in effect scientific — standard of the excavation project. Seventeen 14C dates are many, but given the given the problematic calibration curve probably a hundred would have been necessary.

Despite the time depth of the hearths it seems that use and maintenance didn’t change their form or their centre. If this kind of pattern survives for hundreds of years one would expect that knowledge of the position of hearths was kept alive. This need not be a great problem if, e.g., the hearths were used regularly. But the fact that they kept their centre and circular form suggests strict regulation supported by a habit of marking out and maintaining the hearth periphery e.g. with kerbstones. All in all, it stands out as reasonable to ask whether all this long-term structure was combined with measured formalism rather than just structural presence.

It is difficult to measure the distance between the individual hearths because we cannot know exactly where their centers were, but if we measure the extremes of the sections and calculate the average distance we may succeed. There are five sections consisting of 4 + 8 + (12 = 11+1 empty space) +12 + 4 hearths. In addition there are some pairs in the east end of the row.

Ultuna Fig 05 and 06

From the ‘hole’ in the third section, i.e. a missing hearth, t can be seen that one of the 12 hearths provided for was never constrructed. From the anomaly in the fourth section – a forgotten hearth module – it is obvious that owing a miscalculation two hearths had to be fitted into the space of one in order to obtain twelve hearths. It can be seen that befittingly this is done by dividing the available space into three rather than two. Thus we should expect only 11 modules in this section and we understand that the mistake in measurement was made when defining the position of the extreme hearths suggesting that the space of a section was defined before the hearths were filled in. The fault was detected, but impossible to correct because the surrounding hearths were already in place.

Ultuna fig 07

Ultuna fig 07A

Ultuna Fig 07BCalculating the module, weighted in relation to the length of the sections, suggests a length of c. 3.8m and that is the same as 12 foot or indeed a measure that much later was called a stång, i.e. a rod or pole. A rod divided into 12 parts seems also to have come in handy when the mistake in section 4 was corrected. The two hearths were evenly distributed within two modules – 8 feet apart. The anomalies in sections three and four suggest that the hearths in each quartet were set up from the East to the West. It seem particularly significant that for some formal reason no hearth east of the easternmost hearths in section Four could be established. The identity of that hearth and the four next to it could in other words not be changed. One might say that it was already defined as the First and the ones next to it as the Second, Third and Fourth.

Ultuna fig 08

The layout of the evenly distributed hearths in the row at Ultuna is clearly governed by formalism. In principle irreparable mistakes were made because a hearth can be dug down into the natural subsoil only once. In practice repairing at least one mistake was done with reference to formal measurement and we may therefore ask ourselves why it was important to make four hearths formally spaced in a row and why it was important to return to them and keep track of their precise position?

A definite answer is difficult to come by, but that has never stopped archaeologists from speculating and why should it? Theory after all is the speculative way of pointing out an understanding that may later on be supported or refuted by the contextuality of that which is observed on the basis of theory.

Since returning to the hearths was a habit we may suggest that the space between them represented time as well as module. One hearths could therefore be the first in the row followed by the others all together representing a series such as solstice – equinox – solstice – equinox. In due time when there is a tendency for the hearths to form pairs they may be solstice – solstice or equinox – equinox. What happens in the row is a representation of the cyclic in the linear.

If a community needs to emphasize a time period, such as the year, in this way, why must there be a series of hearths quartets and not just four? And why are there small hearths rather than the remains of large bonfires?

If the point is to gather around the fire in some sort of community the scale of the hearth quartets fits a situation in which each farm or family – constituting a (family) line in history – would gather around its hearth in half circle, i.e. at a seasonal station, e.g. facing south.

Ultuna Fig 09

Since the Late Bronze Age LBA coast at Ultuna is a south coast we can imagine a 12 or 6 o’clock fire in late December, March, June and September. The hearth represents the family’s source of warmth and light and the seven original row quartets seven families on the LBA peninsula. The additional quartets and pairs represent the changing system and changing family structures.

This kind of reconstruction is obviously speculative — a fit of formalism — and far from the practicalities of constructing the row. As any theoretical construction it nevertheless serves as a model for further investigations of other rows of hearths.


(1) On the platau in the calibration curve, see Fig 3 in the text at:  http://www.lcm.rug.nl/lcm/teksten/teksten_uk/a_high_chronology_uk.htm.

This week on the Reading Rest I have the second report from the excavations at ’Paul Young’s Pen’(cf. May 28 2012). This time on a well-dated site created between the 11th and the 3rd century BCE.

Eklund, Susanna, Lindkvist Ann and Wikborg, Jonas. 2012. Påljungshage – kremerat, paketerat och respekterat. Ett gravfält från yngre bronsålder-äldre förromersk järnålder. SAU Rapport 2010:11. Uppsala. Societas Archaeologica Upsaliensis. Pp 356 (1).

For decades Swedish archaeology has had difficulties coming to terms with the grave concept. Discussions started in the late 1960s on formal source critical issues nourished by doubts about atefact combinations in graves mirroring a certain point in time, i.e. the funeral date. During the 70s and 80s the importance of contextualizing graves became obvious, and by the 90s the ‘naïve hypothesis’ became a household term among the critics of traditional grave archaeology. It was used when they wanted to point fingers at these archaeologists, their self-fulfilling prophecies and anachronisms. Graves as a proxy for all kinds of social matters and conventions were questioned, and contextualizing the traditionally unimportant became the important. The traditionalists didn’t bother and the in wake of post modernism they thrived and allowed themselves to avoid the complications of the empirical turn in the 2000s.

Today, in the 10s, as well as during the last decades, it’s within contract archaeology that old and new attitudes to graves oppose each other in a series of texts more informed by practice, tradition and chance than by theoretical considerations (2). This is especially true among the scholarly low-skilled archaeologists in the County Administrative Boards (CABs) and  their guidelines. Thus, in order not to bite the hand that feeds you, the stated goals of the excavation plans given in to the CABs are still heavily marked by traditional views. But now and then the actual results will frustrate the planned goals in a subtle and stimulating way.

The site at ’Paul Young’s Pen’ is a case in point. It was excavated as a cemetery – the CAB would not have allowed any redefinition of or fuzzy theoretical discussion about that concept — but close reading the report, not least between the lines, the ridge at Paul Young’s Pen stands out as a Late Bronze Age (LBA) arena where the past, commemoration and afterlife were cherished – i.e. deposited, nourished and preserved. It was a place for spirits, less complicated than many others and thus easier to grasp. When reading the report the average CAB will mistake the site for a cemetery.

Sometime in the beginning of the Bronze Age (BA) when, after a mellinnium, the water level started to drop the from 20 odd metres above sea level (20 masl), the Stone Age island with the cosy protected beaches (On the Reading Rest May 28 2012) disappeared and became a peninsular. Three hundred years later, when the sea level reached c. 15 masl, the peninsular was a prominent rock or small ridge in the local landscape. As Ann Lindkvist (AL) shows in her analysis of visibility, the site and the top of the ridge was one of several prominent points in the landscape behind and above a subsistence area, a farm land, to the east and southeast. Here settlements may well have been located. Already in the EBA it was decided to build a monument on the top of the bare rock – a very BA thing to do.

The event started with a bonfire. The remains would soon have disappeared, but in a small shallow cleft a little soot happened to be preserved until a monument was built. The monument was a low round stone setting delimited by kerbstones and filled with gravel and earth. When this new ‘top cap’  was ready a central pit was dug in and filled with burnt, cleaned and crushed bones as well as a little coal from the cremation and ritual treatment of a human body, c. 127 grs of bones, less that 10 % of the expected. Later on this procedure was repeated a little to the south southwest of the center, c. 87 grs, less than 5% of the expected weight, went into this deposit. From a number of cremated and crushed bones, c. 25 grs from one or more human beings, scattered in the southwest quadrant of the monument, we learn that the monument now and then attracted remains from cremations.

At Paul Young’s Pen, combining the sacrality of the rock top with a stone setting, BA man created a suitable place where local spirits, represented by a part of their cremated remains, could dwell in afterlife.

The two pit depositions with cremation remains are 14C dated (3).


If there is no shortage of money on the contractor’s part (i.e. if time is valuable) and if a team of archaeologists tell a CAB that they want date the graves they hope to excavate, then the CAB will probably allow a large number of 14C dates, given that artefact combinations are not to be expected. Burnt bones are common and they can be dated quite well if we think they are LBA — a period in which the calibration curve drops steeply. In the transition LBA-PRIA, nevertheless, the curve has a well-known plateau prolonging the calendar year, i.e. the  sister of the 14C year. Dates are thus not very sharp even if their standard deviation is no more than 25 14C years. Notwithstanding, 14C dates are the best.

All depositions of human bones at Paul Young’s Pen are 14C-dated and these dates give us an overview of the use of the site during the last two millennia BCE. There’s an obvious beginning and end to the c. 800 years when monuments were built and human remains deposited. There’s a central period too, covering 100-150 years centering on the 7th c. BCE. Compared to at the calibration curve, the central period reminds one of the time period itself, which means that the 7th c. was a period of frequentl depositions. Since the trend of the calibration curve is falling while the trend line of the depositions is raising, we can infer that the depositions were more common in the end of the 7th c. than in the beginning. Around the central period there are century-long buffers with but a few depositions, and in the 10th and 4th c. there were none.

It stands to reason that the top monument was created to inaugurate a potent place in nature by fire, monument and a deposition of human bones. Later on when depositions began again, the top monument became the root of a more communal place. Eventually the depositions stopped. In the 3th c. BCE someone recognized the place, and put a modern rectangular stone-setting end to the site making it a recognized ancient monument — a former spirit site tagged with a geometrical monument typical of its day and age.

Chronologically, the two depositions in the top monument are wide apart, and it would have been interesting to date the scattered bones in its southwest quadrant to see if they would have filled the gap and enhance the spiritual biography of the monument. Feel free to convince the CAB that another five or so 14C tests are needed, if you think you can.

In the end, the site consisted of 36 deposition in 56 monuments in stone, two (perhaps three) activity areas and the odd hearth. The contextual situation is neatly summarized and discussed by Susanna Eklund (SK). Depositions and monuments are often combined, but one deposit is without a visible monument. Nevertheless, the deposition defines the centre of a possible monument that was never built. One monument was heaped around a bolder.

More importantly, 24 stone settings are without depositions. As pointed out by Jonas Wikborg (JW) a deposition, whether in a monument or not, may traditionally be called a grave. As it happens this is true even though the buried remains, often less than 50 grs of cremated human bones, are extremely pars pro toto. If the pars is no more than c. 100 gram, then the totus of an adult would allow a spirit to be deposited or spread at 25 odd places.

Were such places ‘the grave’ of someone or twenty five? Were people ‘buried’ next to themselves, as well as strategically in the subsistence area?

Or did they, in afterlife, simply inhabit the space where they once lived? And/or were they, as spirits, rooted in their community on the ridge?

It is not difficult, therefore, to agree with JW that monuments without depositions are not someone’s grave in our sense of the word although they may well be memorials. As it happens, a complete LBA or PRIA deposition site is unthinkable without positively ‘empty’ monuments.

Looking at the distribution of the monuments, two of them stand out because they occupy the primary and secondary top point and because they alone are situated directly on the bed rock. The rest of the monuments have a tendency to cluster next to the bare rock. Given this distribution, the site can be divided in two parts along a north-south line. The southeastern part, in its turn, may be divided in two, albeit less obvious parts.

If we look at the chronology of this tripartite site, each part was opened up as one in a series during a hundred year period (indicated by the arrows in the below diagram). The southwestern and the central parts refer to the prime top monument and the northeastern to the secondary top monument. In the beginning of the deposition period the southern part is the most popular, later on, the central and the northern part grow in popularity.

It quite possible that the depositions belong to local settlements and settlers who think of themselves as clusters related to a specific beginning represented by a deposited spirit. The interpretation of the place as spiritual is enhanced when the activity areas are taken into consideration. They are stone constructions abutting the monuments, a heap of fire cracked stones, a cultural layer and a hearth in the northern part of the site with no apparent connection to the bone depositions. These constructions are related to fire and the use of stones. It is possible, moreover, that there was a site specifically use to burn flint in the slope 60 metres southeast of the of depositions. This site is contaminated by medieval activities, but a bronze pin and ceramics link it in time and place also to the depositions.

The contents, composition and construction of monument and deposition vary within narrow limits, and there are indications that deposit as well as monument may be visited and somewhat changed over a time period. If we are into grave archaeology this is problematic because traditionally it is important for the chronology of the past that the dead are buried only once, in a way that reflect the time of their death – not constructed as remains and spirits during a period of afterlife.

Monuments 662 and 792 form an instructive pair that highlights this dilemma. They are old and contemporary – the second and third deposition, in 14C terms, situated next to each other and similarly constructed. The bones were deposited in an urn at the center of each monument, and the urn surrounded by stones that supported a capstone above the urn. Removing these lids gave immediate access to the urns. In the top of both, there was resin and a fragment of a bronze neck ring – in all probability, in each grave a fragment of the same ring (JW p.58f.).

One of the questions that the project was able to add to the original ones was the dating of bones as well as resin rings when found together in the depositions. Resin, used in bentwood vessels to cover the seams, can be quite old because it may be reused in a new vessel when the old one is discarded. Thus, from a source critical point of view it is interesting to compare resin dates to bone dates. Resin from the three old bone depositions — the second, third and fourth deposition according to 14C — was thus dated parallel to the bones.

When the resin was obviously part of the bentwood container (the ring in the bottom of deposit 5988) the dates were overlapping, but in monuments 662 and 792 the resin in the top of the urns was younger than the bones! In fact the resin was the same age as the second bone deposition in the top monument. This indicates that during the central period of deposition there was a need to revisit the earliest monuments and bone deposits. If it was meaningful to add a small bentwood box to the contents of the urns, then the neck ring fragments could have been added too – comforting the spirits of the dead in their afterlife – effectively sabotaging a traditional archaeological presumption.

From a landscape point of view the northern part of the deposition site constitutes an arena with a heap of fire cracked stones in front of a number of monuments related to the bare rock. The secondary top monument, the heap of fire cracked stones and the flint burning site align, and thus the spirit site ‘Paul Young’s Pen’ consists of a southwest-northeast axis following the ridge and a northwest southeast axis following the slope. The site opens towards the southeast and was probably meant to be approached from that direction. In the end the solitary PRIA stone setting marks the end of the first axis along which the deposition site tended to expand.

Talking to all the spirits in their monuments on the ridge is easily done from the flint burning focus at the bottom of the slope. From their dwellings on the ridge the spirits will form an audience looking at the flint exploding in the fire. Indeed the explosions may bring them to attention and some may smile attractively.


The spirit site at Paul Young’s Pen is an afterlife dwelling site. It belongs to one of those periods in Scandinavian prehistory when spirits were not confined to graves, coffins or chambers — when unsepulchered they enjoyed each other’s company. As everyone else they lived in society and had their places contextualizing the fact that there’s a great theological point in knowing whether a spirit must be confined to a grave, e.g. waiting for resurrection, or whether it will live on in an afterlife rooted in the world of the living, even if the spirit is also dispatched from the pyre to the next world. It would seem that an active unconfined afterlife is essential to the LBA, but also that building society on religious faith is human and sometimes bizarre.


(1) The title translates: ’Paul Young’s Pen’ – cremated, packed and respected. A cemetery from Late Bronze Age-Early Pre Roam Iron Age and the short summary states the following:

During the summer and autumn of 2007 Societas Archaeologica Upsaliensis (SAU) conducted an archaeological evaluation and an archaeological investigation of the monument Helgona 220, in Helgona parish, Södermanland. The investigation was necessitated by Nyköping Municipal Council’s plans to build a shopping centre on the site. The investigated site was located just outside Nyköping. Prior to the evaluation, a solitary tone setting on a natural rise in the terrain was known on the site. The evaluation showed that there were late Bronze Age cremations in the stone setting, as well as a large number of other stone settings adjacent to it. The excavated features consisted of 57 stone settings, 1 boulder grave, and 1 unmarked grave. In total, 36 burials were found, all of them cremations. Using radiocarbon and artefacts, the graves have been dated between the Late Bronze Age and the Early Pre-roman Iron Age. Finds consisted of burnt bones, ceramics and resin, as well a small number of bronze and iron artefacts. Also, the burial ground contained a cultural layer, a heap of fire cracked stone and a hearth. In the southeast slope below the actual burial ground, a Late Bronze Age clasp was found, as well as ceramics and burnt flint, all within a confined space, limited by shallow ditches. This area is presumably linked to the burial ground. The slope also harboured a hearth and two postholes.

(2) Those who want to look further into these matters can read Anders Kaliff: Fire, Water, Heaven and Earth. Stockholm. Riksantikvarieämbetet 2007.

(3) One of the defence lines of traditional grave archaeology has slowly become a mantra:

Iron Age 14C-dates of graves are not worthwhile because artefact combinations give better dates
Iron Age 14C-dates of graves are not worthwhile
14C-dates of graves are not worthwhile.

The purpose of an opinion turned mantra is obviously not to convey a simple fact, but rather to concentrate on not-being-bothered by change.