Danish Ring Fortresses



Scholars imagine that the Danish ring fortresses were close to unassailable in their time. They are strong and strategically located, quite obviously according to a centrally conceived plan. They are not only military strongholds but also constitutes examples of technical craft feats. In addition to the tangible work effort, mathematical and engineering skills and profound organizational skills are shown. Today we know seven ring fortresses or so-called trelleborgar, five of which are in Denmark. These are situated from the north like this: Aggersborg on the narrow eastern part of the Limfjord and then south at Fyrkat at the inner end of the Mariagerfjord. Then a little bigger jump past Jellinge, the old Danish royal seat, to Nonnebakken in / beside Odense on the island of Funen. In the middle of the south-western coast of Zealand next to Slagelse, we find Trelleborg, which was first excavated of all the ring fortresses, begun in 1934. On the east coast of Zealand at Køge (just south of Copenhagen) in 2014, Vallø Borgring was discovered, the fifth ring castle. This is, of course, not yet properly investigated.

In the now Swedish part of the former Danish realm (Scania/Skåne), under the present-day town of Trelleborg lies a ring castle, which was investigated and partly built up in the last few years, and a little north of Lund, as late as 1998, a second Skåne ring castle, Borgby was found near Borgby Castle. This ring fortress seems, in contrast to all the rest, to have been used/inhabited even in the Middle Ages. A coin weight and a coin from the 1000’s have been found and in the beginning of the 1100’s a stone tower in the middle of it was erected. Aggersborg has an external wall diameter of 288 m, Trelleborg in Zealand, Vallø Borgring and Borgby about 150 m, while Trelleborg in Skåne, Nonnebakken and Fyrkat have about 120 m.

The most remarkable thing about these ring fortresses (except for Borgby) is perhaps their short lifetime, about ten years and this in the light of all the horrible work effort to build them. One of the things that put archaeologists on the tracks of Trelleborg in Zealand, is a local ”fairy tale” of a lost warrior here. But locally the ring wall has always been visible. It was even photographed from the air in the early 30´s and a local motorbike club wanted to make a motocross track there; this prompted the excavations proper.

The ring fortresses, are narrowly dated, that is to say, all but the newly discovered Vallø Borgring at Køge near the Sound have the same building date of 980/981. For example Vallø Borgring hardly presents anything to the naked view, it has been discovered through advanced laser measurements of the landscape as well as traditional archaeological sampling methods. Impressive ring fortresses in fairly good state of today are Aggersborg, Fyrkat and Trelleborg in Zealand.

There is a different, completely contemporary, construction about 10 kms south of Jellinge: a 760 m long bridge structure over a swampy area, which saved a battle army two day marches in detour. This well-constructed wooden bridge can be perceived in the same context as the ring fortresses, the more as the bridge’s usage time must be set for a maximum of 10 years. There are no signs of repairs to it, and wood is rapidly deteriorating in a marshy environment, unless it is oxygen-free. The bridge structure has then simply been left for deterioration.

What was then the Danish realm?

About the year 980 (or more generally the late 10th century) the Danish realm had begun to strengthen again. The Northenn peoples had sailed across Europe’s rivers and seas for over 100 years. The vast trade network had received goods moving in different directions and caused small towns and trading venues to grow up. Possibly a road network could also have been built up. (This last is probably a little speculative. We know the bridge at Jellinge and we know the so-called Hærveien [Army road], which is pre-Viking age and mentioned in early sources, ie. along the eastern Jutland water-deviding line, which offered comfortable land travel without troublesome river valleys and swamps. Probably there were at least cleared paths here. This Hærveien will run by Hedeby, Jellinge, Fyrkat and Aggersborg. To say anything more of road networks is probably unsafe, although Denmark’s general geography can be said to be well-suited for roads. The most important mode of transport was by sea.)

But politically, the realm of the Danes was severed in the early 900’s after being stronger during the 9th century. There were local chieftains who became powerful enough to be able to covet the rule throughout the Danish area. One such was Gorm the old, who has been counted as ”the first king of Denmark”. This is to say, as far as we know, that he ruled only over Jutland, probably through a patchwork of associated local nobles. Gorm was rather ”the first among equals” (primus inter pares). So-called kingly power was not hereditary, but the sons of the king would then have to fight for it. Gorm, dead in 958, according to an interpretation of the rune stones at Jellinge, is said to be buried in one of the two Jellinge mounds and his queen Thyra in the other. However, this is unclear.

Gorm’s son Harald Blåtand (Bluetooth) should then have been aware of his insecure position and seems to have been aiming at eliminating rivals. The highborns had their own armies and fleets and must be kept an eye on. When one of Harald’s grandsons returned from a foreign journey and began to spend a lot of time with the nobles, Harald is said to have let murder him in an ambush. Blåtand seems to have added the Danish islands alongside Skåne to his realm of alliances. But even Harald’s position was threatened. South of the border beyond Danevirke, built far earlier in the 8th century, for example, Emperor Otto threatened him with weapons. He must have looked longingly at the Danish riches, that is to say the fertility of the country, but for the rest of Christianity, his statements about redeeming their souls from their pagan faith should have seemed laudable.

But Harald went ahead of him by taking the baptism, which he recorded on the greater of the Jelling stones: ”The Harald who won all Denmark and Norway, and made the Danes christian”. The word ’Christian’ is best understood as the thinnest surface veneer for now. In addition, he made an extra reinforcement to Danevirke (according to a tradition, Gorm’s still surviving queen Thyra would have done this). But Otto nevertheless struck through Danevirke in 974 and occupied the trading town of Hedeby. The population, who still worshiped the Nordic gods (according to the Arab Al-Tartushi), was probably not so fond of the baptism of Blåtand, especially when it turned out to be in vain. (In Denmark, pagan funerals have been found well into the 1000’s.) Harald needed to keep his grip on his empire.

Ring Castle Strategy

If you try to imagine a ring castle, hypothetically, near Jellinge a little north of Funen on the east coast of Jutland, you can clearly see the semi-circle of ring fortresses through the inner parts of the Danish area. (A ring castle at Jellinge has not yet been found, but recent excavations have unearthed a strong and wide palissade around the Jellinge area.) We will later describe the structure of the ring fortresses, but already we mention that within them were four long houses in 4 squares (with a square open yard in the middle), ie. in total 16 long houses. Depending on the long house length, which in turn was due to the inner diameter of the fortress, it has been estimated that the ring fortresses could accommodate 500-800 warriors each, which at that time is a considerable force. (Aggersborg would have had 48 long houses and close to 5000 warriors). Each castle will then be at about 30-40 kilometers distance, a normal day march in let’s call it an easily walked terrain (with or without road / path). Vallø Borgring is located at Køge Ådal, which during the Viking era was a sailing fjord, from where one could reach the Scania parts of the realm within one day.

Now let’s look at the locations of the ring fortresses. They are all well away from the outer borders of the Danish area and seem not to have guarded against external enemies. They are close to coasts, but retreated from the sea itself and in connection with this only through a narrow waterway. They, therefore, are not in themselves natural harbour places and they are closed to marine attacks through the narrowness of their watercourses and at the same time not given starting points for marine assaults. They, however, offer waterborne transport of food or warriors. We can take the case of Fyrkat furthest into Mariager fjord. It was well protected by partly marshland and a river. From the fortress through the swamp there was a buried canal to the fjord. As for Trelleborg in Zealand, the castle is again partly surrounded by swamp and in the confluence of two rather windy river branches, pouring 3 km away into Stora Bält, a strategic watershed. However, the river only permits arrival by small ships.

Aggersborg is located on the north side of the narrow eastern part of the Limfjord, protected by a front island and at the crossing of Hærveien and Limfjorden. This, of course, is a very strategic place, far into the country, but this time with full access for larger ships. There is no big reason to linger at the ring fortresses covered by contemporary city centers and their original possible strategic conditions, such as the Nonnebacken, in principle, covered by contemporary Odense and Skåne Trelleborg covered by the Swedish contemporary city of the same name, in addition to considering the possibility of that the modern cities could have originated as small commodities / trading venues under protection of these fortresses. We still have to deal with Borgby northwest of Lund in Skåne. Borgby is located at a small river a few kilometers from its outlet in Öresund, the same picture as for Fyrkat and Trelleborg in Zealand.

We have the currently understood narrow period for the ring-fortresses’ building, namely 980-981. Previously, the intended dating stretched further, and thus the historical understanding was another. Harald’s successor Svein Forkbeard conquered England in 1013 England, which he held briefly, to be followed up by the Danish Canute the Great, whose North Sea empire lasted a little longer. The Ring fortresses’ structure and location do not match the purpose of establishing camps for warriors for maritime conquest purposes. During Harald’s time there were no such war campaigns out of Danish territory. And the building time is under Harald’s time. The purpose must have been another, most probably to keep the population and its aristocracy at bay.

Harald must have had wide views. Consider the enormous work effort behind which there is a clear will for centralization and power. He wanted to create a kingdom in a more modern sense with geographical boundaries. However, he lived in a time when loyalty ties between people still were the norm and their fluidity and nature did not allow strict geographical boundaries. The price, both economically and in measure of confidence, for the building of these ringfortresses may have been high and we will now discuss the structure of these.

The structure of the ring fortresses

The fortresses are built according to a strict common pattern. It is not true that some people built their own kind of ring castle here and others their kind of ring castle there, instead there is a clear management plan. The fortresses consist of perfect circular earthen walls with 17 m width and 5 m height (slightly varying per ring fortress). There are 4 ports, basically straight to the north, to the south, to the west and to the east. This applies regardless of what is outside: a river, marshland or land. If there were solid land outside, there were also moats outside. You can walk straight through the ring castle (in the present time when the ports are missing) in the direction north-south and west-east without encountering houses / remnants of houses. Probably there were wooden streets corresponding to this.

In the midst of these 4 circle quadrants each had 4 long houses to form a square open yard. Some extra house/s/ could also be found. Within Aggersborg’s circular wall each quadrant accommodates 3 groups of these 4 long houses. Between the house groups and the inside of the walls there is space for a circular road. The moats were of about the same width as the walls and 3-4 m deep (in Aggersborg’s case 1,5 m). Aggersborg’s moat almost circumreaches the fortress. Scholars do not think this moat was filled with water but instead studded with pointed poles.

Trelleborg has some differences as compared to other ring fortresses. In front of the outside part with solid ground there is an outlying wall partition concentric with the main wall with a number of long houses with their short sides towards the inside of this exterior. Their longitudinal axes meet in the center of the ring fortress. We speak literally about correct geometry in all respects. The house lengths are exactly the same for each ring castle, somewhat varying per castle and in Trelleborg’s case, depending on which long houses, the outer long houses were slightly shorter than those that lay inside the castle itself. Regardless of the exact length measure used, the Roman foot (about 29 cm) or the so-called Trelleborg ell (49,3 cm), all houses in the ring fortresses have even multiples of this length in their measurements. One has also wondered if specifically Trelleborg would have had a similar wall from an earlier stage about contemporary with the oldest Danevirke. This is due inter alia to some common features of these structures. The last extension, however, is from 980/981.

We now go into features of the structure of the ringwalls themselves. They consisted not only of sloping earthmasses as of today, but they were propped by an inner transverse posture of fixed horizontal and vertical posts to hold together the earthmasses, where there are also stones. The exterior of the fortresses has been covered by a vertical oak tree palisade, connected with the inner posture and secured with external obliquets. At the top of the wall’s outer part was a standing nipple parapet attached to the standing postage. Ditto a lower palissade on the inside of the castle walls. Notice that we are not talking about stone walls. During the Viking era, no stonewalls were built in the Nordic countries, and there was still no warfare in the surrounding area, which demanded stone defenses.

It should be noted that ring fortresses are unique to the Danish area. There are also Viking town walls in other parts of the Nordic region, often semi-circular with aperture toward water and sea. These places, of course, have a different function than the ring forts, who serve the power to hold together a kingdom, primarily landbased. T

he presence of warriors does not mean that only warriors have lived and trained in the ring camps. There have been found traces of crafts, including smith’s craft (not only iron for weapons but also gold and silversmithy) and in the burial fields of ring forts there have also been found skeletons of children and women. From Fyrkat has been found a grave with a völva or seeress with her seeress staff. Despite the fact that ship accessibility to, for example, Trelleborg, was limited, they have found a lot of iron nails and wood that seem to be connected to ship building or ship maintenance.

We proceed to describe the amounts of soil filling resp. timber that was needed to build a ring castle with all its long houses. In the case of fillings for the actual walls, surely some will come from the excavated moat locally, but this is not enough at all. First, the moat is not as deep as the walls are high, and the moat usually covers only one / lesser / part of the perimeter of the castle. A fort like Trelleborg has been estimated to have needed more than 20,000 cubic meters of earthmasses to its ringwall, which can be estimated to a freight requirement corresponding to 1,600 lorries. As for oaktimber requirements an estimation is about half of all oaks of the entire Zealand. If you convert this into transport you get a stunning mileage.

Aggersborg has demanded almost four times as much of earth and gravel masses. If we talk about timber harvesting, it should be noted that the ring forts possibly have been planted near forests of large oak trees, which have produced all oakwood. For a single house in Aggersborg it has been estimated about 66 large oaks according to calculations, i.e. about 5000 large oaks for all houses. We are probably talking about deforestation of oak forests widely around. Why just oak wood? It is the most enduring wood, that is why.

Who were the workers?

An attempt has been made to build a single copy of the old long houses, at Fyrkat and at Trelleborg. Each roof was covered by about 25000 roof shingles, which ”someone” has to manufacture. From these modern attempts, it has been concluded that it has taken about 50 full-time employed men 2.5 years to build a single wooden house of the Trelleborg type. The raw material of oak must be in place at the house. What was required in labour to build the fortress itself with its various details is best left to speculation. To complete an entire working ring castle with all its houses in just two years, including cutting and transport of oaks, excavation and placement of moat graves, conversion of these soil masses to the actual ring walls, retrieval of further large earthmasses and the construction of the posts and the palisades, would have required an extremely significant workforce. This has to be housed and fed during the construction of the building.

So who were the workers?

Part of the work was mostly of the type of wear and tear, but other work was rather qualified as probably the posts in the earthmasses, the palisades and the gates. Ditto applies to the long-house buildings, these were slightly arched as to both height and length, highest and respectively widest in the middle, lower resp. narrower towards the house ends. Craftsmanship is required to get this together. One could imagine thralls who did the raw job, while the qualified work could have been carried out by a local army group, who was faithful to Harald. One should also consider the place-name of Trelleborg, in which the word thrall, Swe. träl (slave), is very likely a part. We should pay particular attention to the Skåne city of Trelleborg, which is known to have at least early medieval origins.

The skilled work could probably also have been carried out by hired farmers from the surrounding area, but these may have been too busy with their own livestock, their own harvest and maintenance of their own small longhouse to be available in larger numbers. Regarding the long posts of the long houses, they were put down in post holes in the ground, with some draining gravel bed. The groundwater attacked these supporting posts, which had to be replaced after about 20-25 years. Due to the better ground they were not as vulnerable as Harald’s bridge posts across the swamps south of Jellinge. The technique of isolating a wooden house from the soil moisture through some type of stone foundation only arose during the Middle Ages.

We simply do not know from the sources conditions for the workers. A small note by the 12th-century chronicle writer Svend Aggesen can support a hypothesis about the use of warriors. Aggesen claims that the king would have sent an army to move an extremely large stone. This could be a distorted memory of the fact that people were commanded for a large construction work, whose meaning they did not fully understand. Another hypothesis is that the king would have used craftsmen from Vendland, that is to say the Slavic populations from the southern Baltic Sea coast from somewhat east of Denmark and eastwards. The reason for this would be that from the eigth century onward there are similar, more simple, wall forts of similar type to Danish ring forts, which seem to be an advanced further development. Similar simpler wall forts are also found in Frisland. Otherwise there are no further role models in northwestern Europe.

Denmark has, according to various sources (including the Old Norse literature), had connections to Vendland and archaeologically, there are some vendic-type finds on the southernmost Danish islands, which can indicate some Vendic / Slavic elements in the population.

Warriors of the ring fortresses and what happened

As mentioned above, the families of warriors may also have lived in the ring camps. The warriors seem to belong to some kind of elite and seem to have been well-fed and wealthy. One sees it from grave items in the grave fields of the ring forts. At Trelleborg has been found a silver cup, glass beads and thin gold plates, something archaeologically characteristic of (pre-christian) grave goods for the higher social classes. Skeleton-wise, they were well-grown young men whose teeth are free from tooth decay, which otherwise was common at the time. One has investigated the presence of chemical trace elements in teeth from these warriors. The exact levels of different trace elements will shift between different districts and their groundwater contents during the growth of the individual. In any case, some of these warriors seem to have come from Vendland.

Such a descent suggests that Harald was in need of warriors who lacked relatives in the vicinity of the ring fort (possibly all ring forts) and therefore only had loyalty to him. One might ask how to feed wealthy and perhaps foreign warriors in larger numbers, who did not devote themselves to the basic farming and livestock farming industry? One possibility is that the warriors collected debt or tax in natural products from the ring forts’ surroundings to themselves and to Harald. The idea is perfectly reasonable from a purely medieval point of view. But now this is still a late Iron Age culture, where there were covenants and loyalty with their own through gifts and return gifts / return services. Tax was something that was taken from other people, not from their own people or their own allies.

Perhaps you can assume that Harald Blåtand and what the ring fortresses stood for were deeply hated by the common Danes and their other chiefs?

In 983, Harald’s son, Sven Tveskägg (Svein Forkbeard), had reconquered Hedeby and Danevirke. His star among the Danes had thus risen, to which may have contributed that he was still ”old-fashioned” in religion, at least according to Saxo Grammaticus i.e he still worshipped the old Norse gods. He rebelled in the year 986 (about 5 years after the building of the ring fortresses) against his father’s rule, probably somewhere in Jutland, and his father fled, mortally injured, aboard a ship to Vendland, where he expected protection but soon died. This according to a number of different sources. Here again there is Harald’s connection to Vendland. Harald was unable to create a geographic realm in a later Christian or national sense, he tensioned his loyalty bonds in a world which held them absolutely reciprocal until they burst and he himself was killed.

The great silence

As explained above, the construction and existence of the Danish ring fortresses must doubtless be regarded as the most remarkable feat of its time. These forts should have given echo again and again in later chronicles. Although politically speaking, the ring forts were abandoned within twenty years after their building, the ruins must have been visible and startling for at least one hundred years afterwards. Human memory and the oral tradition should have preserved their memory into later chronicles. Though it is only mentioned, as noted above, that Harald would have commanded his army to move a large stone block according to Aggesen’s chronicle. Saxo Grammatic mentions the same event. (The block would have been intended to embellish his mother’s grave, Queen Thyra. The warriors are said to have become tired of Harald for this trailing.) Otherwise, there is full silence about the existence of ring forts.

The king we are talking about, who made the ring forts, is, however, ”Denmark’s first Christian king”. Normally, later Christian writers carefully account for the actions and tales of such kings. The Norwegian Christian kings Olaf Tryggvason and Olaf the Holy have raised a whole early medieval literature and even ”Sweden’s first Christian King”, Olof Skötkonung, is quite well-documented in literary sources. (Old Norse preserved text sources do not focus on Sweden.) But not a sound about the Danish ring forts. There is a stark contrast between the heavy presence of the ring fortresses archeologically and their equally significant literary absence. You are tempted to think of a conscious suppression. Now, we are talking about something that happened at the end of the Viking era, near the verge of writing, not something from a dim and forgotten 600’s.

Nevertheless, there is something in the Old Norse literature, which seems to be an echo of the Danish ring forts, and it is about the jomsvikings Jómsborg. This Jomsborg is, according to various sources, to be located to (noticeably!) Vendic area, to Wollin at the Oder estuary. We are not talking of ring forts or fortifications in Denmark itself. There is a proven archeological Scandinavian presence during this time at Wollin, though not of a kind that implies dominance. Jómsvíkinga saga (written in the late 1100s, that is, when the crusades to the ”holy land” had begun) depicts that the Jomsvikings in their Jomsborg would have been their own political unity at the level of Danish or Slavic kings, just as the crusaders were in their time.

Jomsborg is portrayed as a stone castle on the seashore with a gate that could be opened and let in guest ships inside the castle; after negotiation with the Jomsvikings and their leaders. The jomsviking chieftain is designed as an absolute ruler through the strict warrior laws that prevailed there. These are reminiscent of the rules of order in a crusader order. What has just been said about Jomsborg is, of course, an anachronistic fantasy story and no such archaeological Jomsborg is existant. Nevertheless, the wounded Harald fled to Vendland. Adam of Bremen says that Harald fled on a ship to the Slavic city, called Jumne and died there from his wounds. Saxo says that Harald fled to ”Jullin”. According to the Ágrip af sögu Danakonunga (story of kings of the Danes), Harald was wounded by his own son and ”flýði til Jómsborgar í Vindland”, escaped to Jomsborg in Vendland. And an independent source, Encomium Emmae Reginae, says that Harald was defeated by his son and had to flee wounded to the Slavs.

It is clear that Harald had friendly relations with Jumne / Wollin, but also that there were Vends / Slavics who ruled here, not any specific Jomsvikings or even Danes. A deviant voice about Wollin is, however, Svend Aggesen’s chronicle, which seems to indicate that it would be Harald Blåtand, who would have founded Wollin.

The Jomsvik saga also tells us that Danes and Jomsvikings together would have drawn to Norway and lost the battle 986 at Hjörungavåg, which is considered a historic battle. Two named Jomsborg leaders, Sigvaldi and Búi, are also mentioned in other Norwegian descriptions of this kind, but not as Jomsvikings but as Danes. In addition, there are scattered tales in the sagas of the Norse kings that single Vends would have been in Norway. Vends and Norse people from Norway are not near neighbours, nor do they speak the same language.

There are few other possibilities than that the Jomsvikinga saga’s (and some other Norse texts’) Jomsborg has a distorted reality background and as model has the Danish Ring Forts. The actual reality of the ”jomsvikings” were then the Vendian warriors who, according to dental analysis, at least partially populated these Danish ring camps. They can certainly have had special rules of camps and Saxo Grammaticus discusses generally special army regulations in Denmark. However, hardly any ”rules of Order”, as stated by the jomsviking story.

If we are to designate a single ring castle as a model for the Jomsborg fairy tale, then we need to highlight Aggersborg. This is the northernmost ring castle in view of the ”jomsviking” participation in the battle of the Norwegian Hjörungavåg. Aggersborg could have served Danish interests in southern Norway. Aggersborg differs from other ring fortresses. It is many times larger than the rest, it is the northernmost and it seems to have had full access to larger ships through its location. It would therefore be more maritimely offensive than other ring forts. These appear defensively oriented and concern rule over land.

We could easily have developed many more themes. That the Danish Harald Blåtand (Blue Tooth) really had land interest in southern Norway at this time, is apparent from his rune stone: ”The Harald who won all Denmark and Norway and made the Danes Christian”. The same is true of Snorri Sturluson’s Heimkringla or the Sagas of the Norse kings. The purpose of this paper is, however, merely to illustrate what the Danish ring fortresses involved locally. We simply stop with this finding and leave the major politics of that time, along with its later historical (Old Norse) interpretations.


Facts and Fancy in Jómsvík saga, by Leszek P. Słupecki (Polish Academy of Sciences, Institute of Archeology and Ethnology), Saga Conference 2006

Illustrerad vetenskap och historia, a special edition: Yearbook 2014 from January 2015

Jómsvíkinga saga in suitable translation, online, try www.heimskringla.no

Saxo Grammaticus Danmark’s Krønike – Saxonis Grammatici Historia Danica was printed in Paris 1514 by Kristjern Pedersen. Translated by Fr. Winkel Horn to modern Danish from Latin. Available to search online.

Vikinger i krig (Vikings in War), by Kim Hjardar and Vegard Vike, Spartacus Publisher, 2011. ISBN 978-82-430-0475-7