The Past Uniquified – this Time a Boatyard in Southern Sweden
4 February, 2013
This Week On the Reading Rest I have something very common, i.e. a report (again in Swedish, but with good pictures) concerning a ‘unique archaeological’ find. Presently, entering ‘archaeological’ returns c. 43 and ‘unique archaeological’ c. 21 million hits on Google, and among the latter the ironic usage is probably ‘quite unique’, i.e. ‘like uncommon more or less’. Nevertheless, the combination, “quite unique” + ‘archaeological’, returns 400,000 hits. Luckily, when scrutinized the archaeologically unique turns out to be trapped in contextual meaning and entangled in the commonplace.
Nordqvist, Bengt. 2011. Våtmarksfynd från the forna åbädden vid Grönån [Wetland Finds from the Ancient River Bed by Grönån (i.e. Green River)]. UV rapport 2011:11. Arkeologisk undersökning. Swedish National Heritage Board.
Select: UV Rapport 2011:124.
In Scandinavian discussions within maritime archaeology sails have always been important because they indicate technological progress and refinement (cf. OtRR 25 July 2011). As it happens maritime archaeology is still obsessed with the importance of rational technology and invention as the emblem of historical development, as well as the emblem of modern scientific maritime method. The sail and its consequence the sailing ship are technological inventions that changed society because the sail was the first event in a chain triggering the next: Sail → Sailing Ships → Vikings → End of Prehistory →…. Because development is supposed to be dynamic and because we already know that the end of prehistory didn’t happen until 1025-50 CE, sails mustn’t be introduced too early or History will run ahead of itself.
Despite our archaeological endeavor, there are but a few contexts dating the introduction of the sail. The sailing ship with its buried crew of warriors from Salme on Saarema has nevertheless been dated to the beginning of the 8th c. CE and when published in more detail it will set Vikings sailing the seas and getting killed purposeless for generations unable to get started with the ‘Viking Age’, which began only 793 CE, a year that used to be just after the invention of the ‘Viking ship’ (1).
Since mainstream archaeologists have a gut feeling that technological solutions will become popular only when they fulfill a purpose, they are usually not bothered by technological implications. They simply wait for new sources to come to light dating the introduction of the sail. When that happens, i.e. when introduction is redated, it means that theory-based interpretation has been corrected by simple empirical observations, and mainstream archaeologists, who are always prepared to find something by chance, cry: ‘Eureka!’ – and why not ‘unique!’ while they are at it. Accordingly, mainstream archaeologists are prepared to cry havoc when theory threatens to inform empirical observation, and urge the archaeological observer to cast doubt upon the obvious in the observed.
The wetland finds from Grönån are referred to by Bengt Nordqvist (BN) as unique, which means that there is a good chance that they are actually trapped in contextual meaning and entangled in the commonplace.
Grönån is a tributary to Göta Älv, which is the most important waterway in Southwest Sweden. It connects backwoods and coastland. Together with its tributaries, such as Grönån, it also runs through and structures a number of fertile settlement areas. During the Iron Age (IA) the most strategic settlement areas along this river would seem to be situated on the left shore of the river just north of its bifurcation at Kungälv. Sitting here, able to subside between coast and inland, one could benefit from exploiting the hinterland as well as from incoming and outgoing traffic without the immediate risk of being descended upon. Although the area is thus inland, up the river and relatively protected, choosing a settlement by Grönån rather than directly by Göta Älv would suit the general tendency for Early Iron Age (EIA) settlements, contrary to Late Iron Age (LIA) settlement, to avoid the relatively speaking unprotected ‘coastal situation’ by the large river.
On the left side of Göta Älv between Lödöse (the Early Medieval town directly by Göta Älv) and Nol we find a suitable upstream settlement area characterized not least by its maritime finds preserved in the wet and clayey sediments along the rivers. The ships from Äskekärr, a boatyard from the Carolingian Iron Age (CIA, 750-1025 CE) and consequently situated directly by Göta Älv are the most well-known. After Bn’s excavations we may add a similar, but much older site by Grönån. I will call it Skepplanda 226 (Sk226) because the parish name Skepplanda is significant and because it is conceivable that there are other similar sites along Grönån. The name Skepplanda is composed of the words land (in the plural) and skip—ship or skipvidh, i.e. the lands or district that supplies boat timber. BN, doesn’t comment on Skepplanda, but shows that waste and wood chippings indicates that trunks were brought to Sk226 to be processed, making radially split planks (boards), oars, ribs, oarlocks, etc. In addition there are several tools, such as wedges and a mallet, as well as boat details that support the boatyard interpretation. Sk226 is a site that befits a skipvidh-land, but also a landing-place that belongs to one or more farms on higher grounds the odd kilometer north of the river. Situated at Grönån, rather than Göta Älv, the site is protected and comparatively peripheral, i.e. typical EIA compared to CIA Äskekärr. Grönån, contrary to Göta Älv, is an obvious artery for floating small amounts of timber to yards near its estuary. As it happens Äskekärr is situated just 3km downstream from the point where Grönån falls into Göta Älv making it easy to supply the boatyard at Äskekärr with Skepplanda timber.
The reason why we find sites such as Äskekärr and Sk226 is a combination of the need in prehistory to maintain ships and boats, and the fact that rivers because of sedimentation and erosion change their course within the riverbeds. This leads to sediments covering some of the old shores preserving wood and fibers once dropped there by the boat builders. Today the river fronts are protected areas usually not touched by contract archaeology, but in the Skepplanda case the new road E45 had to cross Grönån, and since the span of the bridge was made as short as possible, the edges of the stream had to be investigated.
The excavation was difficult, but very successful and thanks to the sedimentation the stratigraphy revealed two chronological phases divided by sediments deposited c. 375 CE when the river moved c.15 meter to the South. Above the sediments the site continued to be used to land boats. In order to do so despite the new circumstances characterized by a wet strip between the river and the dry land, short canals were dug making it easier to pull the vessels out of the water. BN argues convincingly that in this phase the boats and ships that went into the canal had keels and a cross-section that we usually recognize from ships belonging to the CIA – that is sailing ships. Since the upper Skepplanda stratum is dated after c. 375 and before the 600s it would seem that sails were introduced sometime during these centuries. Salme and Sk226 thus make it likely that sails, but not the Viking Age, were introduced well before 800 CE. The fact that the canals were dug to fit the cross-section of the vessels indicates that the boats were floating when they entered their canal and that the canals were similar to a dock yard from which the boats were probably dragged up on the dry land.
In the first phase the site was used much more intensively in connection with the maintenance of boats. The site was established in the 2nd c. CE, but the hundred years c. 260-360 CE were its heyday. It stands to reason that Sk226 was not meant mainly to give service to people punting and rowing up and down Grönån. Using the boats on Göta Älv and in the archipelago or for coastal trafic in South Scandinavia would seem more important – not least why the popularity of Sk226 in the 3rd and 4th centuries, i.e. the LRIA, coincides with an economic boom in tandem with a period of warfare in South Scandinavian. In the 5th, 6th and 7th centuries we may instead expect lower economic activities. Skepplanda circumstances must to some extent have brought about the boom, but places such as Sk226 must nevertheless boom if the boom is going to be as general as it was. Sk226 is a site that fits the types of boats one would have built in the first half of the first millennium CE. A long, narrow and light boat such as the one from Hjortspring or a long, broad and heavy ship such as the one from Gokstad would not have fit the yard. But a series of boats similar to the ones from Nydam (the pine boat) or Valsgärde might well have been maintained at Sk226 (2).
The report is an overview with an understandable, but nevertheless biased focus on the ‘unique’, i.e. so far in Sweden the oldest wooden artefacts of their kind, definition, character, and function. Thus influenced by the face value of the ‘unique’, the report lacks the systematic information that would have allowed the reader to interpret the character of the site, i.e. its commonplace context, in greater detail (3). Consequently, the report doesn’t discuss intra-site matters in relation to time or the three periods defined by the 14C-tests.
Two examples go to illustrate this: (1) Although part of the canals belonging to the upper layers cut down into the glacial clay below the lower layers, we don’t know precisely where these disturbances of the lower layers occur. (2) The chronology rests on 14C-tests, but it is difficult and sometimes impossible to find out where the dated artefacts were found. Despite shortcomings such as these, there are nevertheless clues here and there in the text; some conclusion can probably be drawn and some insights sifted from it. Despite the restricted attitude in the report to conveying contextual facts, contextualizing the site is worth a try.
If the lower artefact layer at Sk226 is a boatyard used with varying intensity during 2-300 years, then one may wonder, especially looking at the last 100 years of the LRIA, whether the find distribution reflects the structure of a yard. The hypothesis (i.e. a theoretically informed point of view that may be tested against empirical patterns sometimes overlooked) is the following: maintenance takes place when a boat is lifted out of the water and secured with props allowing the craftsmen to stand next to the boat working on. In this position, while working, they tend to lose tools and utensils and to drop waste and broken parts. We would expect these wooden objects to be trodden down into the humid clay by chance and perhaps eventually by intension making the ground more stable. Since there are planks that must be wedged out of trunks, dressed and cut by the ax, we can expect wooden chippings and bits of planks more or less all over the place, but not specifically under or next to the boats. This means that (1) where the boats stood finds ought relatively speaking to be few. (2) Next to the boats there would be concentrations of artefacts; and (3) between the boats there should be more chippings and pieces of plank than next to or under the boats. These hypotheses hold true only if some parts of the site are defined by their boat standings. In the long run they are unlikely to hold true given the general freedom to choose a boat standing as best you please.
In the report the artefacts are divided into groups and mapped. Insignificant pieces of wood were plentiful but not recorded. Wooden chippings were abundant and fragments of planks relatively common. Both categories were recorded and mapped because being marked by tools they might reveal something about craftsmanship. If we use the categories defined by BN and combine the artefact maps in the report to give an impression of density and complexity there emerges a pattern to some extent consistent with the hypotheses indicating where the boats stood when they were maintained and repaired.
The pattern is easy to see in the two peripheral distributions, but blurred in the dense central part of the site where the boats may have stood within a larger less specific area just west of the landing-place. The peripheral boat standings are characterized by relatively speaking fewer artefacts.
When we add the chippings and the planks to the artefact distribution it is enhanced inasmuch as the areas between the possible boat standings are filled up by signatures. Since wedging the trunks, and dressing and cutting the planks benefit from an open work space rather than a boat standing, the two distributions match each other. Obviously axes were used next to the boats too, but less intensely.
If we figure out where the four oldest, i.e. the 2nd century 14C-tests were found their distribution coincides with the dense central part indicating that over the centuries the ideal division at the center of the site – the landing and the standing – becomes blurred because of the activities going on. It seem fair therefore to conclude that in the centre of the distribution, next to the landing area, later marked by the canals, there were standings where for hundreds of years boats stood when they were not used. In the period after c. 375 CE when the boats had their standing above the landing-place, maintenance might well have taken place at Sk226, but in that case the dry conditions of this possible site has deprived us of wooden remains. Nevertheless, the analysis of time and space at Sk226 suggests that the central part of the site was used between the 1st and the 7th c. CE. In the LRIA, work in the yard was intensified and this intensification shows in the addition two or three peripheral standings left and right of the original site. Because the peripheral standings were not used during a longer period and because they were left undisturbed by additional landings, their find distribution is not blurred. It tells us that from time to time in the LRIA there was more than one boat being looked after at Sk226. The expansion indicated by the 14C-dates is in other words matched by a find distribution that suggests a growing demand for tonnage in a period of dynamic economy and warfare. Before the activities at the site come to an end they became less intensive and thus more similar to the activities of the ERIA.
Grönån is a small river and the boats and ships that landed at Sk226 in the RIA were hardly more than 12-15 m. Probably they were similar in size to the small LRIA (pine) boat from the war offerings at Nydam. Oarlocks and a number of oars suggest that they were rowing vessels. A helm indicates that it wasn’t just small boats that landed. Some of the finds parallel elements of the LRIA boats in Nydam and one detail, a block with no wheel and thus perhaps a kind of gutter ring, might fit a rig. Since part of an anchor was also found the object might perhaps have filled a function in that connection uniting rope and anchor. Because of the Nydam parallels, the boats probably had a cross-section similar to the large Nydam boat and according to the new reconstruction of this boat the hull must have been supported by props when the boat stood on its standing in the yard. Probably the weight of such a boat was c. one ton. If we imagine that boats were now and then to be dragged past waterfalls, and compare them to what we know about the capacity of the large Nydam boat, they had a crew of c. 18 giving each man 55-60 kg to drag. Speculating about the reasons for giving up Sk226 it is reasonable to point to the fact that if we are engaged in regional and inter regional transportation the ships that could be maintained in Grönån became too small. A sites such as Äskekärr would accommodate large boats and ships although it lacked natural protection and easy conditions for floating timber. Sk226 was probably too small-scale and too protected for the new dynamic times of the CIA.
Sk226, Äskekärr and Fribrödreå (on Falster in Denmark) make up a small series comprising a millennium of off-settlement South Scandinavian ship yards in wetland environments characterized by wood because iron is not preserved. Their complement, linked to settlements such as Lundeborg (on Fyn) and Parviken (on Gotland) and covering the same period, are the dry land yards characterized by iron rivets because wood is not preserved. Together these sites define an economic geography and it is comforting to know that just as they are few and far apart today, they were once commonplace.
(1) Salme http://www.academia.edu/244134/Rescue_excavations_of_a_Vendel_era_boat-grave_in_Salme_Saaremaa
There is an abstract: Warrior Burials with Scandinavian Finds of the Late Vendel Period (ca 750 AD) from Salme in Saaremaa/Ösel (Estonia)
Jüri Peets, Raili Allmäe, Liina Maldre (Tallinn University, Estonia) and Ragnar Saage (University of Tartu, Estonia)
In autumn 2008 remains of human skeletons and ancient artefacts, including some deformed sword fragments, boat rivets and two antler dice, were brought to light while digging an electrical cable trench for the lighting of a cycling track. Deciding mainly by the shape of the weapon fragments they were dated to the Vendel Period or the beginning of the Viking Age (7th–8th centuries). The finds were of Scandinavian types. Some of them, including gaming pieces and single-edged swords, hadn’t been previously found in Estonia. Artefacts related only to Estonia or Saaremaa were missing. The excavations of the site were resumed in 2010 and 2011, revealing a second ship – a big warship about 17m long. The ships contained the skeletons of 43 warriors with weapons and other grave goods: about 40 swords, 12 shields, about 50 arrowheads, 12 horn combs and about 300 game pieces lated of whalebone and 20 bovine femur heads, 5 of them were ornamented. Alongside with humans, dogs and hawks were sacrified. As food offerings swine, goat/lamb and bovine bones were found. The most significant find beside the ancient artefacts were the discovery of the remains of the first prehistoric boats (ships) in Estonia.
And information in Estonian on http://www.delfi.ee/teemalehed/juri-peets
(2) Nydam: http://www.abc.se/~pa/uwa/nydam-e.htm or Rieck, Fleming. 1994. Jernalderkrigernes skibe. Nye og gamle udgravninger i Nydam mose. Vikingeskibshallen i Roskilde
Valsgärde, the boat in grave 6 looks like this:
(3) Seemingly, the County Administrative Board (CAB) didn’t ask for it. At page 74 the author refers to some questions, posed before the final excavation took place and they may well have been formulated by a CAB because they are so odd:
Why is the largest artefact group objects that can be connected with the production of planks? The whole process of production from trunk to plank is represented in the material. Do these remains emanate from the production of something specific – such as boats?
If planks are what you find and boats what spring to you mind then you obviously know something that has already provoked the affirmative. The following question is thus not really surprising: Do even abandoned boats occur? But then again you better ask yourself because you wouldn’t want to miss them just because it hadn’t occurred to you that there might be boats, do you?
The next on the other hand is puzzling:
Is the agglomeration of the processed wooden objects a natural deposition? One might think that there is no need to talk of natural or unnatural depositions when we may talk of what is presumably meant namely primary or secondary depositions, but then again perhaps not; you never know with CABs. The right answer may actually be: ‘No, it’s an unnatural deposition!’ because you are not allowed to reformulate the original question, are you?