Supported by a grant from the King Gustavus Adolphus VI Foundation For Swedish Culture, osteologist Lena Nilsson has analysed the bones we collected during excavations last year at two Medieval strongholds. Two weeks with 19 fieldworkers at Birgittas udde produced only 0.4 kg of bones, because the site has no culture layers to speak of and the sandy ground has been unkind. But from the following two weeks at Skällvik Castle we brought home 32.7 kg of bones! And now Lena has looked at them all. Here are her reports:
The reports are in Swedish, but the species names and anatomical terms are given in Latin. Birgittas udde was occupied briefly in the 1270s but then seems to have been vacant, though kept in repair long into the following century. Skällvik Castle was occupied from 1330 to 1356 or shortly thereafter.
Lena is available for more work, and I’ll be happy to help readers get in contact with this seasoned osteologist.
Update 29 May: And here’s Lena’s report on the bones from Landsjö Castle 2015.
Our second week at Skällvik Castle proved a continued small-finds bonanza, and we also documented some pretty interesting stratigraphy.
More of everything in Building IV. In addition to more coins of Magnus Eriksson, dice and stoneware drinking vessels, we also found a lot of points for crossbow bolts. It’s starting to look like the castle guards’ day room! As for why we found crossbow bolts only inside one building and none outdoors in the bailey, I figure that they had been amassed there for re-fletching. The dark indoors find context and the undamaged sharp points show that the bolts did not end up in the floor layer of Building IV because people were shooting there.
Building IX has a lovely floor sequence. First a terracing layer with few finds, probably pre-dating the entire building. Then a culture layer. Then a stone cobble floor. Then clay full of magical fairy stones. Then mortar. Then large square unglazed brick tiles, only a few of which survived. Then a demolition layer mixed with household refuse. Such attention paid over time to the floor of a low-ceilinged cellar-like ground-floor room demonstrates the resources available to the castle’s owner.
Building X has had a nice ribbed brick portal, as evidenced by broken decorative bricks left by the quarrymen.
The gate house is still very rich in bones, as remarked on by the 1902 excavator. And his coffee-loving workmen or their contemporaries left a small midden in it after the end of fieldwork.
Two finds in particular document the presence of the social elite, which is hardly surprising at a royal castle where both King Magnus and the Bishop of Linköping dated letters. One is part of a wheel-turned ivory ear scoop, the Q-Tip of the era’s nobility from the NW building. The other is a seal matrix with a simple coat of arms, found near the castle dock. I’m optimistic that specialists will be able to read the inscription and identify the owner.
We have enough coins to be able to draw chronological conclusions not only from which types are there, but from which ones are missing. All but one of the coins we have identified so far were struck for Magnus Eriksson, most during his final minting period about 1360. His successor and nephew Albrecht was crowned King of Sweden in February of 1364, and this ruler is represented by only one coin, a frontal crowned-face bracteate.
We can see the quarrying of the castle for building material, targeted largely towards bricks, preferentially towards specialised decorative ones and floor bricks. These were in all likelihood taken by boat the two kilometres to Stegeborg Castle when it was re-erected, probably in the 1360s. And when Stegeborg was partly torn down about 1700, the bricks travelled on to the royal castle in my home town Stockholm!
It would be great to learn what’s under the rubble that fills the castle keep. But excavating this building would be an enormous undertaking, with regard both to the sheer volume of rubble with large heavy boulders, and to the long-term conservation commitment once you’ve emptied the structure. It is not a job for one precariously employed university lecturer and his students, working for a few weeks on small grants. The organisation that excavates Skällvik Castle’s keep will have to have solid long-term funding. And it must be willing to put a roof on the structure, restore the masonry and return the keep to daily use as a museum space with offices. Any responsible intervention into a ruin must be done with an eye to the far future, not just to the next tourist season.
The famous royal castle of Stegeborg sits on its island like a cork in the bottleneck of the Slätbaken inlet (see map here). This waterway leads straight to Söderköping, a major Medieval town, and to the mouth of River Storån which would allow an invader to penetrate far into Östergötland Province’s plains belt. The area’s first big piece of public construction was 9th century fortifications intended to guard this entrypoint, in the shape of the Götavirke earthen rampart some ways inland and a wooden barrage at Stegeborg. This barrage was kept up for centuries, and indeed, the castle’s name means “Barrage Stronghold”.
Stegeborg is easily accessible by car and receives many visitors. Few of them however then continue on to the nearest castle ruin, which is oddly enough only 2 km to the SSW: Skällvik Castle. It sits on a rock outcrop on the south shore of the inlet, inside Stegeborg’s line of strategic defence and within sight of it. Why two castles so closely apart?
The written sources are rather murky, partly because they rarely differentiate between a manorial land property and a castle sitting on that property. But following my friend Christian Lovén’s treatment of the evidence, the timeline seems to be roughly this.
C. 1300. King Birger has Stegeborg Castle built.
1318. Stegeborg is besieged and largely torn down in a civil war after King Birger has his rival brothers Erik and Valdemar starved to death in a Nyköping dungeon. Reports the Chronicle of Duke Erik from the victorious party’s perspective, “They broke down that wall so completely / They did not leave one stone on top of another”. King Birger is deposed and exiled.
C. 1330. King Magnus (the son of the murdered Duke Erik) or the regents during his minority refortify the Slätbaken passage, but with a new castle on the hill at Skällvik rather than on Stegeborg Island. About the same time, Skällvik Parish church is built between the two castle sites.
1356. Skällvik is attacked by the forces of Erik Magnusson, rebellious royal pretender and son of King Magnus.
1360s. Realising the obvious, King Albrecht (nephew of King Magnus) returns to the superior strategic position and rebuilds Stegeborg, quarrying the ruins of Skällvik for building material.
So Stegeborg and Skällvik Castles aren’t really two separate castles when seen on a strategic scale. They’re two versions of the same castle that has moved slightly to and fro over its 400-year lifetime, leaving a fossilised mid-14th century version at Skällvik.
Unlike Stegeborg Castle, which was maintained and extended up to the 1690s, nothing was ever built on Skällvik Castle’s foundations. Only recently was a small-scale brickworks established at the foot of the castle hill. In 1902 restoration architect A.W. Lundberg considerately removed a lot of the rubble from the ruins, leaving the culture layers easily available to the excavator everywhere except inside the keep. In the past week, me and my team have been the first archaeologists to take advantage of this state of affairs.
In addition to the high keep, Skällvik Castle has three main buildings arranged around its sloping bailey. They are joined up by short stretches of perimeter wall. We have opened trenches inside all three buildings (two in the long building IX-X) and in the bailey along the wall of building X. In addition we have test-screened and metal detected two big spoil dumps from 1902, and skilled metal detectorists have investigated our trenches and the surface around the foot of the castle hill. A few of our main results so far:
Five of six Medieval coins date from about 1360. The sixth is too corroded (i.e. debased) to allow dating before conserved. Two were minted in nearby Söderköping, one in Kalmar.
All other coins date from 1800 or later.
Building IV, where Lundberg identified a baking oven in 1902, has yielded all five datable Medieval coins, a bone gaming die and a piece of fine burgundy stoneware, probably from Lower Saxony. Drinking and gambling in the warmth of the bakery!
Building IX has yielded a fine cobbled cellar floor and a comb fragment.
The two 1902 spoil dumps are erosion rubble and brick kiln refuse, respectively, and not productive of small finds.
Unlike Birgittas udde, Skällvik Castle offers many preserved bones.
I have demonstrated experimentally (and idiotically) that when clearing brambles, you should wear protective glasses. I was lucky to only get my left cornea nicked a little. Hurt pretty bad and left me barely functional for two days.
Stoner dudes parked by the authorities at your religious abstainer hostel provide much entertainment with their spaced-out antics and conversation. Unless you mind food, cigarettes and bikinis going missing. And nocturnal rearrangement of furniture. And heavy bass at two in the morning.
We spent Thursday afternoon backfilling. As I write this, only trench G remains open, and the guys there expect to finish soon. Here’s some highlights of what we’ve learned during our second week at Birgittas udde.
Trench A in the outer moat demonstrated that the moat had a wide flat bottom, was not very deep and contains no lake sediments. Probably always a dry moat, providing material for the bank behind it. No Medieval finds.
Trench C in the inner moat demonstrated that this moat too had a wide flat bottom, but it was deeper and is full of hard clayey sediment: seems to have held shallow water. No datable Medieval finds.
Trench D inside the inner defensive bank demonstrated that the site has been used for quite some time and rearranged: the stone pavement proved to be the top of a deep fill used to obliterate a rectangular cellar with drystone walls. Apparently this is the cellar of an early building that has been replaced by the large mortared masonry cellar that dominates the site today. It would have saved considerable labour to extend the cellar instead of replacing it, which suggests that this was the stronghold’s main building that couldn’t be torn down until its successor was habitable. Disappointingly, no Medieval finds in or under the stone fill.
Trench F demonstrated the same general trait of the site as trench D: there are two generations of two house foundations here, all in all four. No further Medieval finds after the glass shards and the coin.
Trench G at the northern end of the western line of buildings has given an iron latch lifter (that is, a simple key) and a spread of stones that suggests the building’s northern gable may have consisted of a drystone wall.
Let’s summarise what we’ve learned so far. Birgittas udde has quite a long Medieval use history where at least three buildings in the inner bailey have been replaced with newer versions. The high nobility makes its presence known not only through the fortifications and the large intricate masonry cellar, but also through a broken imported drinking glass. The complete absence of pottery in eight (or nine) indoor and outdoor trenches, though, suggests that waste management was quite fastidious, or that the site was never inhabited for very long each time, or both. This supports an interpretation where Birgittas udde is a special-purpose site for threatening situations, a fortified retreat in a peripheral location on Ulvåsa manor’s land. And the main Medieval version of Ulvåsa manor is at Gamlegården, the long-lived unfortified site abandoned in 1580.
A few decades ago a badger family dug a large sett next to the big cellar and dumped their spoil into it. The level floor left by the restoration architect in 1924 has had a big ugly spoil mound sitting in one corner ever since. At the landowner’s suggestion, and with the County Archaeologist’s blessing, my team rolled the turf off the mound and lifted it out of the cellar bucket by bucket to backfill the abandoned sett. Despite screening and metal-detecting much of the badger spoil, we made no Medieval finds. Another piece of restoration we’ve done is to give the intact piece of corridor vaulting over the cellar’s entrance a much-improved turf cover. This will keep rain from leaching the mortar out of the arch, improving its longevity.
On the 9th we’re moving our headquarters to Stegeborg camp ground & hostel. On the 10th a number of new team members will join to replace people who are only with us at Birgittas udde. And on Monday the 11th we break turf at Skällvik Castle. Stay tuned, Dear Reader!
Ulvåsa in Ekebyborna is a manor near Motala with two known major Medieval elite settlement sites. Excavations in 2002 proved that the unfortified Gamlegården site was established before AD 1100. The fortified Birgittas udde site has seen no archaeological fieldwork since 1924, when the main building’s cellar was emptied and restored. Its date is only known to the extent that almost all moated sites of this kind in Sweden belong to the period 1250-1500. My current book project deals with Östergötland province’s fortified sites of the High/Late Middle Ages, and so I decided to spend two weeks digging at Birgittas udde with my students this summer. I am lucky enough to field a record team this year: 18 hard-working people with more to join later. We broke turf last Monday.
Medieval Ulvåsa has quite rich written sources thanks to an extremely famous 14th century inhabitant: St. Bridget of Sweden. She spent over 20 years here as wife and mother, raising eight children, before becoming a prolific religious author and major political player. My project isn’t focused on her, but good written sources are always a pleasure to work with, and the top-level nobility to which St. Bridget belonged is central to my investigations. Contemporary sources say nothing about where on Ulvåsa’s land the saint lived. My guess right now is that the family stayed mainly at Gamlegården and withdrew to the peripheral stronghold of Birgittas udde only when the political situation demanded it.
Birgittas udde is a long, high narrow promontory into Lake Boren that has been cut off with two straight moat-and-banks. We have opened six trenches.
Trench A is in the fill of the outer moat and aims to seek the date of the site’s abandonment and hopefully household refuse in the bottom layer. Our most interesting find so far here is a redeposited Mesolithic quartz core which came as a pleasant surprise.
Trench B is in the outer bailey, a sizeable featureless area between the outer bank and the inner moat. The trench aimed to find out about Medieval activity in this space, but instead it has given most of the evidence we have for the Mesolithic use of the promontory. It’s a microblade industry utilising high-quality quartz and red quartzite, very similar to assemblages from other sites in similar locations around the lake. We were lucky enough to have three friends of mine, all major authorities on the Mesolithic, visit us and classify our finds: Lars Larsson, Fredrik Molin and Roger Wikell. They place the finds in the Middle or perhaps Late Mesolithic, around 6000 cal BC. We backfilled and re-turfed trench B yesterday. No sunken features were found on the trench floor.
Trench C is in the fill of the inner moat and has a similar aim to that of trench A. Our most interesting find so far here is an intriguing lump of slag that speaks of metalworking on site.
Trench D is in the easternmost of the buildings along the bank in the inner bailey, and aims to study the building’s use. The small finds here have not so far been illuminating. But we have encountered a strange pavement of large selected stones that might suggest the foundation for an immense oven, if it weren’t for the complete absence of charcoal and other signs of burning.
Trench E is in the inner bailey on the open featureless space between the various house foundations. Like trench B, it aimed to find out about Medieval activity, but here we found nothing at all. We almost finished backfilling trench E yesterday.
Trench F is our largest one. It includes some of the interior of the westernmost of the buildings along the bank in the inner bailey, a large sloping outdoor surface, and some of the interior of the southernmost of the buildings along the inner bailey’s western edge. The trench aims to study the use of these buildings and hopefully pick up household refuse. Our best Medieval finds are from trench F: two sherds of a highly ornate imported drinking glass and a silver bracteate coin (so far unidentified). The glass in particular is exactly what we are hoping to find.
I don’t think we will have time to do a trench G.
Before signing off I must mention two people who are extremely important to the success of this campaign: my co-director Ethan Aines, who is absolutely ace, and our landlord, Carl von Essen, who is unbelievably hospitable, helpful, interested and insightful.
One more week at Birgittas udde now. Who knows what we may find?
Myself, Ethan Aines and Mats G. Eriksson are proud to present our report on last year’s fieldwork at Landsjö Castle, Kimstad parish, Östergötland. Lots of goodies there! Construction on the castle seems to have begun between 1250 and 1275, and the site was abandoned halfway through an extension project some 50-75 years later. We also found a Middle Neolithic fishing site and an Early Modern smallholding among the ruins.
I spent last week in Denmark at a friendly, informative and rather unusual conference. The thirteenth Castella Maris Baltici conference (“castles of the Baltic Sea”) was a moveable feast. In five days we slept in three different towns on Zealand and Funen and spent a sum of only two days presenting our research indoors. The rest of the time we rode a bus around the area and looked at castle sites and at fortifications, secular buildings, churches and a monastery in four towns. Our Danish hosts had planned all of this so well that the schedule never broke down. Add to this that the food and accommodation were excellent, and the price very humane, and you will understand that I was very happy with the conference.
This was my second CMB. Last year in May I attended the twelfth one in Lodz, Poland. It’s an excellent education for me as I delve into High Medieval castle studies with my ongoing project about castles in Östergötland.
You might think that within such a specialised field there would be lots of debate at the conference, but actually participants present work that is mainly of local or national relevance. Your audience takes a polite interest in what you’re doing, but nobody presents any results or methods that change the game for everybody else. I imagine that this has to do with written history’s specificity. These scholars aren’t dealing with large generalised prehistoric cultural categories. They’re dealing with specific people and events at specific castle sites. If someone has found out new stuff about the architectural phasing of a certain castle in Lithuania, then this will not change the way someone in south Jutland thinks about her subject much. But every specific case presented, and every site visited, offers a wealth of details that add up to help castle scholars contextualise their work at home.
The presentation that I found the most interesting was Christofer Herrmann’s and Felix Biermann’s about recent fieldwork at Barczewko / Alt-Wartenburg in northern Poland. This wooded area, Warmia, saw a planned colonisation effort sponsored by German lords in the 14th century. Written sources document that a settlement was founded at Barczewko in 1326 and razed to the ground by Lithuanian raiders in 1354. Attracted by a long-known but undated defensive bank-and-moat, my colleagues have now mapped the site with geophys and excavated key buildings. The geophys showed a neatly planned mini-town, with a main street, a town square and a town hall. The cellars are still full of the debris from the fires set by the attackers, on top of the goods stored in the cellars, and a few bodies of murdered inhabitants. Almost a little Pompeii, and very painstakingly excavated. The pottery is dominated by Silesian designs (from the south-west part of modern Poland), giving an idea of whence the colonists came.
Late Iron Age settlements are full of copper alloy objects, making them the preferred site category of metal detectorists. High Medieval castle sites, on the other hand, are quite poor in these often distinctive and informative finds.
The picture above shows all the copper alloy and lead that my team of ~15 found in over two weeks of excavations at Landsjö castle this past July, screening the dirt and using a metal detector in our trenches. Only seven objects! We collected 133 pieces of iron in that time, of which 77% are sadly nails in various states of completeness and thus not terribly informative.
172 is a piece of folded sheet lead. I’m going to ask the conservator to unfold it, because sometimes they hide magic spells inscribed with runes.
173 is a piece of thin embossed foil that broke after we lifted it. It just might be a really debased bracteate coin. Conservation will tell.
174 is a half-pipe fitting that has been riveted onto something, maybe a strap. Hoping for some decoration to show up.
175 is a rose-shaped embossed-sheet dress spangle, a ströning. Love it!
176 is a thick domed sherd, probably from a tripod cooking pot.
177 is an 18th century jacket button.
178 is a cylindrical cap made of thin sheet, probably also from the 18th century. I don’t know what this may have been part of.
I write these lines on the day after we backfilled the last two trenches at Landsjö, packed up our stuff, cleaned the manor house, hugged each other and went our separate ways. It’s an odd feeling to take apart the excavation machine while it still runs. It’s been four fun and successful weeks!
Since my previous entry, written on Monday evening, we’ve done only three days of further excavation. Our main findings, to the extent that I have any comprehensive overview of them at the moment, relate to the culture layers sitting between the natural subsoil and the rubble in trenches F and I. The latter in particular yielded some interesting bits: a spur (yay chivalry), a copper alloy decorative mount shaped like a rose, a large iron key, two knives of which one has a beautifully preserved bone handle, and a single lovingly shaped and smoothed limestone flag that had ended up upside down and cracked on the floor of the basement in trench I. Both trenches also finally gave the pottery I’ve been waiting for, Medieval red wares.
Having dug just inside the perimeter wall we got a good look at it at two points. The northern reach in trench F is 120 cm thick, standing on a base that is 160 cm thick, sturdy work. But in trench I there is barely a perimeter wall at all. It has no widened base and sits partly on yellow sand, partly on sloping bedrock, quite flimsy. This wall along the western edge of trench I (on a stretch between a crevasse and a big glacial erratic boulder) is clearly just part of a building and has not been laid out as part of the sturdy perimeter wall encountered in trenches C and F. This reinforces the picture we’re getting of an unfinished stage set of a castle.
Trench H ended up giving a beautiful posthole packed with re-used mortar-stained rubble. This post-castle feature suggests that much of what we’ve been looking at in the trench has to do with the 18th century smallholder’s activities on the island. A shed or barn is drawn just west of trench H on the 1730 map. We’ve also identified the foundation of his home, just east of trench F at the islet’s highest point: a flat rectangular surface with foundation frame stones visible above the turf at two points in the western and southern walls, and a typical rubble mound from the stove and chimney at the northern wall.
Now I’m going to take a few weeks off, and then post-ex work begins. Stay tuned!
Like Stensö, Landsjö Castle has half of a rare perimeter wall and is known to have been owned by a descendant of Folke Jarl – or rather, by his daughter-in-law, the widow of such a descendant. Last year we found that the high inner bailey has a previously unseen southern wall with a square tower at the east end, and we found five coins of AD 1250-75 in a deep layer that seem likely to date the castle’s construction phase. But unlike Stensö, in three strategically placed trenches we found no trace of the missing bits of the perimeter wall.
This year we started a few days early with half the team at Landsjö, and I write this on Monday night of the second week with full staffing. We’ve opened four sizeable trenches this year, of which G showed conclusively that the steep outer bailey has never had the missing half of its perimeter wall. It seems that the castle started with an L-shaped perimeter wall defending only the west and south sides of the inner bailey, and the north and east sides left unwalled – which worked thanks to steep high drops there. But the outer bailey seems never to have reached a defensible state. With it the castle looked rather impressive from the lakeshore, but if you wanted to get into the outer bailey you could easily bypass the unfinished perimeter wall. All we found in trench G was an 18th century jacket button most likely belonging to the smallholder who lived on the islet at that time.
Trenches F and I (just SSE of F) are in the basement remains of stone buildings built in the inner bailey as part of its western perimeter wall. Both basements are full of rubble, which we are still removing, and both are yielding ample finds of bones and small iron objects. Trench F also shows clear signs of a major fire, with lots of cracked stone in the rubble, with the inside of the walls visibly flaked by the heat, and with a charcoal layer just emerging this afternoon. If we’re lucky, then nobody’s been poking around in that burnt layer since the fire went out.
Trench H has been both a disappointment and a boon. It has not yielded any of the walls we hoped for. No gatehouse. Rather it seems to be on an enormous spoil dump from the digging of the dry moat across the castle islet. But in this spoil we’ve found Early Red Ware, a High Medieval pottery type, and, intriguingly, a sherd of Late Neolithic pottery along with some knapped quartz. Neolithic pottery experts who have seen a photograph agree that this is Malmer type J ware from the final phase of the Battle Axe Culture about 2300 cal BC. Fun and unexpected!
Just inland of the swamp woods and rushes covering the shore of Lake Landsjön, the manor’s overseer Roger Österqvist kindly helped us machine a 50-metre trench along the shore just where the distance to the islet is least. And there, under a metre of Carex peat, we found pointed stakes rammed into the clay at six spots. Most of these are probably from simple jetties and fish traps. But two are thick enough to belong to bridges over to the islet. Radiocarbon and maybe dendrochronology awaits. We also found a beautifully preserved iron ard tip, part of an agricultural implement of likely High Medieval date.
Lotta Feldt of the County Museum told me something extremely interesting Monday. A tight and reliable radiocarbon date for mortar can be had for €700 at the University of Turku. This would allow us to date every major construction event at these castles – because we have been taking mortar samples. I’m definitely looking into that!
Now we have three days of excavation left, plus half a day of backfilling before I return most of the tools to the museum and our team disperses. I say most of the tools, because I have had to buy eleven orange plastic scoops designed for bailing small boats. We use them to catch trowelled soil when we clean between stones. I’ve decided that this is a fine statement of wealth and status. I own more bail scoops than any private individual I know, and I don’t mind telling all and sundry.