39 About Royal Manors from the Late Iron Age in Middle

David Damell

How royal power arose in Sweden and how this affected the creation of the State, has always been the focus of discussion among scientists and laymen. This is certainly because royal power can be connected to some of the most important ancient monument areas in the country such as Gamla , Adelsö hovgard etc. In these areas therc are visible remains of'monuments and cemeteries as well ah remains of terraces, plateaus etc. How these royal seats were formed is at present unknown. Excavations in many of these ancicnt momument areas during the last decade have, however, shown that the royal seats were constructed in a certain pattern. In this regard the excavated royal seats at , Fornsigtuna, Adelsö hovgard and Husby-Enhörna are the only seats addressed in the paper. Each of these seats has strong similarities with one another. It seems possible though to start future excavations of other royal seats based upon thc results described in the paper. Dalid Dameil, Örebro lrins museum, Bor 3i4, S-70i 46Örebro, S»eden.

How royal power arose in our country and out until recently. To seek detailed knowl- how it may have contributed to the formation edge about these special places must now be of the Svea State —Sweden —has always been considered as important as continuing the a question of great interest to scholars as well more general studies. In this way we can find as to the general public. The so-called Väst- out whether the places actually served as göta-school is living proof of this interest. royal manors, and if so when they were estab- The royal power is connected with some lished, how they were designed and used etc. of our most distinguished ancient monuments such as Old Uppsala, Adelsö Royal Manor etc (Ambrosiani 1985).This connection between power and specific places may be studied in historical sources. But also the time before written sources points to this connection, for example the connection between certain place- Gatkla Uppsala names and certain types of ancient monu- MJ DDLE SWEDEN ments such as big mounds (Hyenstrand 1989). Fonnsi na onhhntm Adelsok» The connection between royal power and Mos rnh na big mounds has certainly also been studied by an older generation of scholars. Those schol- ars tried to connect older traditions or mythi- cal stories directly to certain burial places of known king» or chiefs (Lindqvist 1936; Ner- man 1941).Burial mounds, excavated as well as unexcavated ones. have so to speak played a role in research. Any attempt to seek knowl- Fig. L Map of mitldie S» edetr»:itll tlte fi)ttr Moral edge about settlement patterns of probable manors metltionetl in tbe article tnarked. Dra»ing by royal manors has, however, not been carried Emilia Lauer-At idet sstttr.

Current S» odish Aretsaeolosn t ut. I, /VV3 40 David Damell

Fig. 2. The aneient mona»rerrt area at Frrr nsigtunaj ust near the estate . Malr by G Wi»lrerg arrd R Rydén. Drarving by Emilia Lauer-Andersson.

After that it is of course possible to return to Signhild's mound. Then two house-plateaus, a more general view, which means a study of 12 house-terraces, one rectangular stone-set- the royal manors from a perspective of state ting, one cemetery with 10 round stone-set- formation. Let us begin with Fornsigtuna, tings and one with 10 triangular, boat-formed situated at the estate of Signhildsberg in Hå- and round stone-settings. tuna parish, Uppland, near Lake Mälaren and Most of the house-terraces are situated at not far from the old town . Immediate- the 8 m level and in connection with stonecon- ly north-west of the manor there is a large area structions, probably remains of landingstag- covered with ancient remains around a mead- es. In the southern part of the monument-area ow. some remains of a medieval village are visi- The 5 m level shows a shallow lagoon just ble. north of this area. The monuments can be The whole area was mapped by the Central described as follows: Board of National Antiquities in 1983—1985. One mound 27 m in diameter and 2—3 m There are also other cemeteries within the high with a flat top. The mound is called boundaries of the estate Signhildsberg.

Crrrrent S» edrsh Arehaeoiogs, Vol. I, IW3 Abaut Royal Manors 41

It is known that the estate Signhildsberg in house-plateaus larger trenches were excavat- older times, before 1678 was named Fornsig- ed. tuna (Old Sigtuna). The name Fornsigtuna is The mound was first believed to be a first mentioned about 1170. In a letter from motte, but the walls found in the top of the the pope Alexander to the Swedish king and mound were very late, the remains of a sum- his earl (Knut Eriksson and Birger Brosa) the merhouse visible on a map from 1817. latter are told to give back three estates to the Under the walls there was a cairn of a type archbishop, taken from him when the diocese common in burials. But there was no burial was moved from the town of Sigtuna to Upp- there, only the remains of a dark cultural layer sala in the 1160s. without objects and dated to the early Roman It is probable that the king in connection Iron Age by C14. Also a ramp in connection with this took back property which earlier had with the mound gave early datings, in that been donated to the church during the mission case to the pre-Roman Iron Age. era, perhaps by King (Hult 1989:49ff). So it was proved that the mound was not Fornsigtuna is also mentioned by Snorre Stur- originally built as a burial mound or a motte. lasson (about 1220) as the place where What I now think is that Signhild's mound first settled in Sweden and where the pagan was originally erected as a court-mound, due cult was practised. Also in 1299 Fornsigtuna to the similitaries with known court-mounds is mentioned in a letter by the Swedish king such as one (Domarehögen) at Old Uppsala Birger Magnusson. It is also mentioned in and other parallels. Court-mounds were used 1315 in connection with the royal family. as judge's seats in very old times. It is proba- It seems quite clear from all this that Forn- ble, though not fully proven, that this mound sigtuna was a royal estate, originally a so- was built in connection with the house-pla- called Uppsala öd-farm. Also later Fornsigtu- teaus mentioned above. The excavation at the na is mentioned in assessments from the 16-th smaller house-plateau, ca 30 m long, was century. done by means of two big trenches, one in the In 1627 Fornsigtuna was donated by the middle of the plateau and one in the eastern Swedish king Gustav II Adolf to a nobleman. part. In 1678 a new manor-house was finished at The excavation has shown that the plateau

Fornsigtuna and from that time the estate had is bui l t artificial ly of a big heap of stone on top a new name, Signhildsberg, from an old Nor- of a small hill. The plateau was then covered dic tale about Signhild and Habor. with clay, sand and soil which composed the All this means that it is a fact that the estate floor in a big long-house 9 x 25 m in size, an Signhildsberg in ancient times was Fornsigtu- area of 225 m-'. There were two rows of post- na, the precursor of Sigtuna (Dame11 1989:22ff; holes and the walls were slightly curved, a v Friesen 1923; Malmberg 1923). very typical construction during the The aims of the archaeological excava- Period. The artifacts found are nails along the tions at Fornsigtuna were to make a map of the wall s, arrowheads (not the hunting type), pot- place, to date the remains and to investigate sherds, a fragment of a bronze ornament, the continuity of the settlement and the func- glass beads, bone material etc. The dating is to tions of the place. As mentioned above, the the Viking Period. mapping was done 1983-1985. Also phos- There are also seven C14-datings from phate-mapping was done. After that, excava- AD 490+ 75 to AD 950+ 190, and one from tions began in 1984. So far field-work has AD 1420+ 90. Most of the C14-datings show been carried out for 5 months. The field-work the Viking Period, is now finished. Only small trenches were The other house-plateau is contructed in investigated in eight of the house-terraces. In exactly the same way as the one mentioned Signhild's mound as well as in the two big above, but it is much bigger. This house was

Curren( S» eitish A rcluirntng, Vot. 3, 399.3 42 David Damell

40 x 10 m large — a huge building. We know Age, especially the and Viking Peri- that this very dominant building dates from ods. The results of the excavations so far are the Viking Period, judging from the artifacts very good in relation to the aims mentioned and C14- and TL-datings. It is also worth above. mentioning that this large building is situated The examination has shown that there is a near the entrance-road to the whole estate- quite clear continuity within the place at least complex. from the early Iron Age. The two plateau-buildings must have dom- It might be that the excellent conditions inated the whole area, and hence it is clear that for communication at this place made it a they played a special role here during the natural meeting-point for cult and jurisdiction Viking Period. As mentioned above eight house- for a larger area during the early Iron Age. It terraces were also investigated by means of is possible that the court-mound was erected small trenches. In all cases the remains of already at this time, but it is more probable buildings could be observed, with cultural that the mound is younger because of the deposits and artifact material very typical for similarity with the house-plateaus seen from a the late Iron Age. Also from the terraces there constructional point of view. Sometime dur- are C14-datings, the oldest AD 340+ 85, but ing the shift between the Roman Iron Age and mainly within the time span from AD 610 to the Migration Period the first house-terraces AD 955. seems to have been erected. From that time on In 1988 the remains of a rectangular build- they must have been used continuously until ing, 9 x 9 m large was also excavated. It was the 10th century. impossible to date this foundation properly All these results rest mainly on C14-anal- but from the construction it seems quite clear ysis. But also the artifact material is datable that it must belong to the same time as the especially to the Vendel and Viking Periods. plateau-buildings as a part of the whole com- The oldest object from this place is a glass plex. The function of this building also re- bead dated to the Migration Period. C14- mains unclear but it seems probable that it was datings and dated objects seem to point to a a store-house, a very typical building in a more intense activity within this place from complex like this. the Migration Period onwards. This can be The phosphate-mapping was done in a interpreted as an indication that the activities coordinate-system at every 20 m. The sam- at the old court and cult place gradually ex- ples were taken at a depth of 30 cm. The panded to include for example markets. phosphate map shows very clearly that there Quite soon a growing central-power in the is much phosphate in connection with the province of Uppland was established at the house-foundations and also within the area place in order to get influence and control, where the medieval village seems to have above all over the economy. A royal site was been situated. etablished. This must have happened during Also osteological analyses were made. The the when the house-plateaus results show that there are similarities be- were erected. There is so much work behind tween Fornsigtuna and Birka with regard to this establishment that it seems probable that the mixture of different animal bones. A com- there was an intention of supreme power be- plete analysis of the artifact material has not hind it. Also the dominant position in the yet been done. Most of the material seems in terrain must be interpreted in the same direc- fact to be connected with the building con- tion. The two plateaus are so similiar that I structions, such as nails etc. But there are also believe they are the work of the same archi- parts of ornaments, a bronze weight, arrow- tect. My theory is that the royal hall was heads, beads, pottery etc. The material is gen- situated on the smaller plateau and the garri- eral, but it is also very typical for the late Iron son in the hall on the big plateau near the

Curren( Sw erlish ~rc lvie«lusv, Vi)l. I, IVVa Afaout Royal Mattors 43

Fig. 3. The Arlelsir comtrtes nith tlre nuirrs rrf Alsnii hu» (uo 2h burual rnorrnds (no 4—7) etc. A f(er. Thondernan /920. entrance-road to the estate-complex. It is also ruins of Alsnö hus, some other medieval building clear from historical sources that Fornsigtuna remai n s, three burial-mounds (King' s mounds), was once a royal site. This place, then, must a runic inscription on a boulder, a so-called have been used as a royal site until the end of court-mound, the burial mound Skopintull the 10th century, when among else the good and a lot of terraces, probably man-made, to harbour conditions just north of Fornsigtuna the east and north-east of the Alsnö hus ruins. were silted up by land elevation. This fact and The latter are the remains of King Magnus the necessity of creating a new place for the Ladulås' palace from the later part of the 13th new religion, Christianity, caused the king century, excavated by Bengt Thordeman (- (probably Erik Segersäll) to build a new town deman 1920) who at the time also registered for market and cult ceremonies on land be- some other remains of medieval buildings in longing to the Crown but not far away from the ground to the west and south of Alsnö hus. the old central place. The new town also got In connection with Thordeman's excavations its name from the old place, as a sign that the deposits of context were also old supreme power was now moved to a new found, above all in a terrace straight east of the place. ruin but also just to the west and north of the The old estate was then donated to the big medieval foundation (Rydh 1936:186ff). bishop of Sigtuna in the 1 lth century, but No constructional detail» were found on that when the diocese moved to Uppsala the Crown occasion but the find material, consisting of took back its old property (Ambrosiani 1985; fragments of Arabian coins, beads of glass Damell 1989:28ff). and cornelian, iron arrowheads, parts of a The next important place is Adelsö hov- bone comb and potsherd» is typical Viking gård (Adelsö Royal Manor) on the island of Age. Adelsö in Lake Mälaren. The ancient monu- Direct information about the character and ment area at Adelsö hovgård consists of the size of the Viking Age dwellings was not

Currenr Sreerlrsh Arrrtraeotrrgs, Vol. /. 1993 44 David Damell

Fig. 4. Husby-Enftiirna with building remaitts, grave-mounds ete. Map by 8 Elfstrand. Drawing by Etnilia Lauer-Andersson. possible to get from the very limited excava- Hanna Rydh (Rydh 1936: 104 ffl and situated tion. Topographically the Adelsö complex is within the ancient monument area of the royal very similar to that of Fornsigtuna, situated as manor, underline the importance of the site as it is on the top of a dominating promontory early as the Viking Age. The very rich finds well visible and with a vast view over Lake must be interpreted as remains of royal burials Mälaren and with the adjacent Viking town of datable to the 10th century. Birka within sight. If we go southwest from Adelsö to Over- One hypothesis is that the town of Birka enhörna parish in the province of Söderman- was founded under the protection of Adelsö land we find the remains of another royal Royal Manor, in the same way as Sigtuna was manor which has been partly excavated in founded under the protection of Fornsigtuna recent years, namely Husby-Enhörna at pre- Royal Manor. In order to study some ques- sent-day Ekensberg Manor. The estate got its tions about the origin of the prehistoric royal name in the 17th century after the governor manor, its size and appearance, its relation to Bengt Ekehjelm. He got the estate from the other royal manors and to the town of Birka, State, which had owned it since the so-called new excavations begun at the site last summer Västerås recess of 1527. The older name was (1991).Of course the medieval layers on the Husby. During the 13th century Husby be- site will be analysed at the same time, as will longed to Duke Erik, a son of Birger Jarl, but the connection between the royal manor and later it belongs to the cathedral of Strängnäs. the neighbouring Adelsö Church. It is also Before that Husby must have been royal prop- important to point out that the rich finds from erty. Within the ancient monument area, Hus- the burial mound Skopintull, excavated by by backe, Bengt Elfstrand in collaboration

Current Swedish Arehaeoingy, Vnb I, t993 About Roval Manors 45 with the Husby-Enhörna Foundation excavat- Not until very recently were excavations ed several building remains showing that the carried out on the courtmound and the royal site had been used continuously from the manor terraces under the leadership of Else Viking Period until the 17th century, Nordahl (Christiansson & Nordahl 1989).The The place is very similar to both the earlier trial excavation on the top of the court-mound described places, situated on a former Mälaren showed that it was built in a way very similar bay. A big plateau, probably artificial, domi- to that of the court-mound at Fornsigtuna, nates the site. Around it several house-foun- with a stabil izing stone layer covered by a thin dations were found. Just to the west of this layer of gravel and sand which is covered by complex there are several Iron Age burials clay under a sandy turf layer, the surface of and some more building remains, probably the construction. There are no traces of buri- from the late Iron Age/Medieval Period. A als or other activities in the construction. small harbour must have belonged to the site. From my point of view it is therefore probable Notable among the oldest finds are an amber that this mound originally was built as a plat- bead, a glass bead, potsherds, fragments of form for the court, then as a court-mound. It loom weights, knives, nails etc. has not yet been possible to date the court- The design of Husby-Enhörna is thus very mound. The upper royal manor terrace (pla- similar to that of Adelsö hovgård, with a plateau for house-foundations dominating the landscape in a communication position and well visible from the Lake Mälaren bay at that time. Just like at Adelsö there is a church near- by, Överenhörna Church, in direct connection with the royal manor. Atboth places, as at Old Uppsala, the churches originally must have served as royal manor churches. Fornsigtuna, which was abandoned as a royal manor al- ready during the Viking Age, does not have a church. But the most distinguished royal manor of all was that of Old Uppsala, or rather Uppsala, as the site was called before the name was transferred from the old place to the town Östra Aros when the diocese moved from the cathedral of Old Uppsala to the new town. This happened during the later part of the 13th century. Ancient monuments and finds from Old Uppsala are carefully described by Sune Lindqvist (Lindqvist 1936)and therefore there is no reason to discuss them in detail here. It is well known that the ancient monument area consists of three huge royal mounds, two of them excavated, a big grave-field to the south of the royal mounds, a court-mound and the so-called royal manor terraces north of the Fig. 5. Old Uppsala rvitlr the rrrval rnourrds (A-C), the church (today only the chancel of the old eourt-morrrrd ())), the ehurelr (E) «rrd the royal marzor cathedral remains). terraees (F-G). AJ)er Lindt)rast )936.

C«rrerrr Su ezhslr Arz haezrtnss, Vnt. l, 3993 46 Dnvi d Damell teau) just north of the church was shown to be well as from an architectural one. It seems artificially built already in the excavations by clear that a special design formed the royal Sune Lindqvist. It was built of a 2.5 m thick manors described here, maybe a traditional layer of clay, and on the top of this plateau layout of the state building complexes already there was once a huge building, a hall, at least during the late Iron Age. A certain chronolog- 10 m wide. The length is still unknown. The ical agreement seems to exist, too, even though only dated artifact so far is an iron brooch the time of establishment seems to be certain from the 8th century. only for the buildings at Fornsigtuna, where It is also proven that a probable prehistoric there are Vendel Period datings for the bottom cultural layer can be found within the church- layers of the plateaus. At Old Uppsala the yard and also under the church. It is also foundation period may be older (Migration important to mention that Viking Age boat- Period in the two excavated royal mounds) graves have been excavated in recent years while Adelsö hovgård may very likely be just south of the Old Uppsala vicarage. contemporary with Fornsigtuna. So far Hus- The royal manor area of Old Uppsala also by-Enhörna seems to belong to the Viking shows topographical similarities with the ear- Period. lier described sites, above all with Fornsigtu- Another similarity between Old Uppsala, na and Adelsö hovgård. Arti l icially built house- Adelsö hovgård and Fornsigtuna is that towns plateaus in landscape dominating situations, have been founded nearby, certainly under the presence of so-called court-mounds and the protection of the king. I am thinking of good conditions for communication are sig- Östra Aros (present-day Uppsala), Birka and nificiant features. The latter also applies to Sigtuna. Perhaps it is also a fact that the royal Husby-Enhörna. manors discussed here were established along I have now briefly described four excavat- the important trade routes by a central power ed royal manors in middle Sweden. Of course of increasing strength, which in the course of the places mentioned are only a small part of time created supreme power over smaller king- the total number of prehistoric royal manors. doms and chiefdoms in the region. This may But they are the only sites so far excavated be the first sign of a state formation created by which give special attention to dating and increasing economical and political control design and for this reason they are important over smaller areas which no longer were able models for future studies of the remaining to resist the power of the main king. material from achronological point of view as

c«rren& 5«ecl&ch A r&v««&&l&t&x, '«d. I, 1&vf&a Alzout Roynl Mattors 47 REFERENCE S

Ambrosiani, B. 1985. Aristocratic Graves and Manors Lindqvist. S. 1936. Uppsala Irdgar rrch Ottarslréigerr. in Early Medieval Sweden. Archaelogy and Environ- Stockholm. nrerrt 4. Umeå. Malmberg, E. 1923.Signhildsberg. Håtuna socken, Håbo Christiansson, H. gz Nordahl. E. 1989. Tingshögen and härad. Uppsala län. Ss:ert ska slott oclr herresiiten. Kungsgåtrds-platåerna in Gamla Uppsala. Tors Vol. Stockholm. 22. Nerman, B. 1941.Sr'eriges ril e» ulrpl onrst. Stockholm. Damell, D. 1989. Fornsigtuna. As stamp —fiir en ny Rydh, H. 1936.Förlrisrorisl a rrrrelersril ningar p&i Adelsii. Sigtuna-for skrring. Sigtuna Museer. Stockholm. v. Friesen, O. 1923. Om staden Sigtunas ålder och Thordeman, B. l920. Alsrrri lrus. Ett svenskt medelri ds- uppkomst. Upplanels forrrmirrnesfÄrerrirrgs tidsktnft palatsi sirr I rurstlristrrriskrr sarnuranhang. Stockholm. 37, 1922-23. Uppsala. See also: Hult, B. 1989.Strom. fornsitune et Guazbro. Sigtunastiftets Damell, D. 1991.Fornsigtunrr. En krrrrgsr;rlrrls historia.

biskopsbord. A v s(amp —fii r err ny Sigtuna-forskning. Stiftelsen Upplands-Bro fornlorskning. Tierp. Sigtuna Museer. Hyenstrand, Å. 1989.Sverige 989. Makt och herraviihle I. Stockholm Archaeological Reports nr 24.

Crrrrerrr Sseertistr An:haerrlngs, tkrrl. I, 1993