By Carl O Nordling

It has long been held that a certain island called Björkö ('Birch Island') in Lake Mälar, east of Stockholm, was the destination of the missionary expedition of the monk Ansgar (later bishop of Hamburg) about AD 830. It is well known that the missionaries Unni and Hiltin in the next two centuries were sent to the same place as the one that Ansgar had chosen. Consequently, there would have been a certain continuity of Christian customs in the place from AD 830 to about 1060.

Björkö is situated in the middle of the cultivated area of the Iron Age, in the valley of Lake Mälar. A Christian missionary station in this place would naturally have required the consent of the King who held power over the same area. It therefore seems a little odd that no King of Swedes took the baptism before Olof Skotkonung did it more than 170 years after Ansgar's mission to Birca. It is also from the 11th century that Christian customs begin to show among the archeological finds in the Mälar Valley. Svealand as a whole remained unconverted until 1080 (i.e. 250 years after Ansgar), according to B. and P. Sawyer1.

It seems very odd, indeed, that a succession of pagan Swedish kings (e.g. Emund and his sons Olof and Erik) apparently should have tolerated a missionary station for an alien religion in the middle of the populated area--and at the commercial centre itself, of all places. At the same time, it is hard to explain why the Christian missionaries did not achieve any traceable results in the surroundings of this well-situated station during the first 170 years of its existence. Most astonishing of all is the fact that when King Olof Skotkonung finally took the baptism about 1008, he chose to do it in Husaby in Västergötland, about 260 km from Björkö and Sigtuna, his new capital.

Apparently, the effort to propagate Christianity among the Swedes was an utter failure during the first 170 years, although those in power did not seek to thwart the effort. Christianity even gained a foothold in Västergötland some time before it established itself in the neighborhood of Ansgar's centuries old station. There seems to be something wrong somewhere, the whole story does not make sense. Can we trust that the sources concerning the missionary work by Ansgar, Unni and Hiltin have been interpreted with all the necessary criticism and discretion?

It has been suggested indeed that the Birca of Ansgar may not be identical with the Björkö island in Lake Mälar. As a matter of fact, there are several places in Sweden that were called "Birca", "Byrcke" etc. even as late as in the 16th century. Also in Finland we find a Birkala or Pirkkala (near Tampere) and a Pirkkiö (in the north)2 . These names are apparently derived from the word birk(a).

Rembert (or Rimbert), the collaborator and successor of Ansgar, has left us a report on their missionaray expeditions in his book Vita Anscarii (about 870)3. There he mentions a place of some importance which he calls "Birca". The same name occurs in the Gesta Hammaburgensis ecclesiæ pontificum (about 1074) by Adam of Bremen3. Apparently both these authors regarded the word Birca as a proper name, and this interpretation is still prevalent. The relevant literature deals with a place bearing the supposedly unique name of Birka. But Birka was hardly a proper name at all and certainly not a unique one, as we have seen. The word birk is generallly considered to have been a common noun denoting a town or market place with certain characteristics, perhaps with a legal system of its own4. Some researchers connect the name Birka with this word. Also, a place such as Björkö (originally Bierkö 'birch island') could be called Birca in Latin.

Therefore, the Birca of Rembert and Adam may well have been a place outside the densely populated area in the Mälar Valley. The island of Åland in the northern part of the Baltic is one of the places suggested as a possible location for the missionary station in question. This island would certainly have needed a market place (that is a birk) in the Viking Age--preferably located centrally on the island.

Adam of Bremen did not himself visit the place he calls Birca, nor did he visit Sweden at all. His narrative is based on tales and reports from a number of various informants, such as Bishop Adalvard Jr., Ratolf of Slesvig and King Sven Estridsson. In order to compose his geographic accounts Adam must necessarily have used various sources. It is quite possible that one source told him about one place called Birca while another source depicted another place also called Birca (i.e. birk, Björkö etc.). In such a case Adam could easily have concluded that both narratives dealt with the same place, the unique Birca that existed in his conception.

After carefully scrutinizing every piece of Adam's geographical information, the Dane A.A. Björnbo came to the conclusion that Adam had neither made up a map by himself, nor been in the possession of any map of the lands that he wrote about. Björnbo also stated that a map of the countries and places mentioned in the Gesta could not at the same time represent all of Adam's distances and directions5.

The present idea about "Birca" as the name of a unique and important place is, however, based mainly on the rich archeological finds that have been made on the island of Björkö in Lake Mälar. Since the first finds were made there about 1680, Björkö has been a natural field for archeological excavation. The finds have confirmed the theory that an important, and rather unique, market town existed on the little island from about 750 until about 970. Few other places in Scandinavia have been subjected to archeological studies matching this. It is therefore impossible to know for sure if this market town was the only one of any importance within Sweden and the coastal areas around the northern part of the Baltic.

Furthermore, it is not a matter of course that Ansgar should choose the biggest market place as a base for his missionaray activities among the northern peoples. He may well have been wary enough not to establish himself too near Upsala, the center with the heathen temple. It would probably have meant courting disaster to defy the heathen gods in their own backyard, so to speak.

All this demonstrates that the name "Björkö" and the rich finds on the island so named are not enough to demonstrate beyond reasonable doubt that this was the location of the missionary station visited by Ansgar, Gautbert, Unni and Hiltin, at least in the years 830, 853, 936 and about 1060. There seems to be plenty of good reasons to submit the sources to renewed scrutiny. Let us try.

Rembert

In Chapters 10-11 of his book, Rembert tells us that Ansgar and his suite on their way to the region of the Swedes (partes Sueonum) met with pirates and were forced to abandon their ships. The party continued their journey by land a very long way and crossed the intervening seas in ships (interiacentia maria navigio transeuntes). In this manner they finally reached the port called Birca in "their kingdom" (ad portum regni ipsorum, qui Birca dicitur).

Comment: If Rembert had wanted to say that Ansgar crossed bays, sounds, lakes and creeks, he could have used words like sinus, fretum, æstuarium, fluvius etc. Actually, he uses the word mare which usually means '(salt) sea' or 'ocean'. A strict interpretation of this passage leads to the conclusion that the party probably landed on the island of Öland and crossed the salt sea in order to get to the mainland and thereafter crossed the sea once again. Thus they might very well have ended up on an island in the Baltic. Such an interpretation could explain both the use of the unexpected word mare and the application of its plural, maria. Other interpretations are certainly possible, presuming either that Rembert did not choose his words particularly carefully or that he was attacked by the pirates already before entering the Baltic.

Rembert goes on with telling that Ansgar was well received by "their ruler, named Björn" (a rege eorum, qui Bern vocabatur). From this we cannot conclude more that Björn was a ruler over some district, large or small. (There was still no "Kingdom of Sweden" at the time.) Apart from the rune stone at Rök, Björn is not mentioned anywhere else. Among the local population some were favorably disposed toward the new faith, especially the governor of the market town (præfectus vici). His name was Herigarius and he is said to have built a church on his hereditary estate (in hereditate sua ... ecclesiam fabricavit).

Chapter 13 tells us that Pope Gregorius IV (827-44) appointed Ansgar to be his envoy to the Swedes, the Danes, the Slavs and other peoples living in the northern regions (in aquilonibus partibus).

Comment: If Ansgar was expected to establish contact not only with Swedes, but also with other tribes (e.g. Goths, Slavs, Estonians, Livonians etc.), it seems natural that he should choose a site for his station at some Baltic port rather than on a lake island in an area populated exclusively by Swedes.

In his Chapter 14, Rembert mentions a Bishop Gautbert (alias Simon) who was sent to the Swedes a little later and who built another church "there" (cœpitque ... ecclesiam inibi fabricare). It is not quite clear whether inibi refers to Birca, but the context seems to indicate that it does.

According to Chapter 19, Birca was left without a priest for seven years, probably about 845-852. During this period the town was attacked by Danes, when Herigarius, the merchants and all the inhabitants got into straits. Rembert says that they "took refuge in the 'city' that was near by" (ad civitatem, quæ iuxta erat, confugerunt). The enemy, however, prepared to destroy the "city" (ad diripiendam urbem).

Comment: The mention of a "city" (civitas, urbs) near by the market town (vicus) Birca seems somewhat peculiar. In medieval Latin the word vicus (from Greek oikos 'house') meant a village, 'a settlement of some importance', or a trading town, shipping centre etc., while oppidum and urbs (later also civitas) denoted 'town enclosed by walls', or even fortress, castle, etc6. If Rembert had seen the market town on Björkö, fortified with its citadel on the adjoining hill, the adequate appellation for it would have been oppidum, not vicus. The words civitas and urbs suggest something larger than vicus and oppidum, but on the island of Björkö there was certainly no larger town or fortification where the town-dwellers could seek refuge. Reasonably, Herigarius and the merchants lived in the largest center in the neighborhood.

But suppose that the Birca seen by Rembert was a certain small settlement on Åland (the commercial centre of the island), a place later on called Saltvik (nowadays Kvarnbo). In that case the word vicus would have fitted perfectly. And not only that. About 2 km from Saltvik there is a rocky hill, 40 m high, with remains of an ancient stronghold. Seen from outside at a little distance, this stronghold, called Borge, would probably have looked just like a typical Europeanurbs surrounded by a town wall. It is highly unlikely that a person who had actually seen the redoubt on the hill (25 m high on the lakeward side) sloping down to the market town on Björkö would refer to that as an urbs. The words that lie near at hand for this object are citadellum, castellum or mœnia, with the addition of contagium or confine rather than iuxta. This passage in the Vita Anscarii apparently does not fit at all with the state of things on Björkö. At the same time it is a good description of Saltvik and Borge on Åland.

Finally it is noteworthy that Rembert nowhere mentions Lake Mälar, nor makes use of the word insula or insula parva 'island, islet', which would have been the obvious description of Björkö. This island measured less than one square mile in area at the time and the insular impression on the visitor was unavoidable.

Adam of Bremen

Adam of Bremen in his Gesta Hammaburgensis (about 1074) provides us with at least thirteen different pieces of information bearing on the location of Birca.

Item 1: (Book I, Chapter 60) "Birca is the town of the Goths in the middle of Sueonia not far from the temple ... called Upsala" (Birca est oppidum Gothorum in medio Suevoniæ positum, non longe ab eo templo, quod ... Ubsola dictu).

Comment: Adam's narrative is about 200 years younger than Rembert's, but it is based on hearsay, not on observation. Adam rejects the term vicus and calls Birca oppidum, which may reflect either a factual change or the change of informant or, perhaps, a change of linguistic usage since the time of Rembert. It is noteworthy that this town "in the middle of Sueonia" is said to belong to the Goths. It seems rather unlikely that the Swedes should have admitted the Goths to establish a market town of their own in the middle of the Mälar Valley.

The inhabitants of the island of Gotland were certainly called Goths (gutar). There is a slight possibilty that Adam extended this term to include the inhabitants of Åland as well. Another possibility is that the Swedes had actually accepted a settlement of Goth merchants somewhere in the outskirts of the area dominated by Swedes. If Adam knew that Swedes inhabited the coastal areas on both sides of Åland (see below), he may in fact have regarded Åland as lying "in the middle of Sueonia".

On the other hand, Adam's statement that Birca was situated "not far from Upsala" tallies much better with Björkö than with any place on Åland. (Nevertheless notice what is said in Item 8 below.)

Item 2: (Book I, Chapter 60) "In this place [where Birca is situated], an inlet of the Baltic [...] directed towards the north forms a port for those foreign peoples who live along this sea". (In quo loco sinus quidam eius freti, quod Balticum vel Barbarum dicitur, ad boream vergens portum facit barbaris gentibus, que hoc mari diffusi habitant.)

Comment: Like Rembert, Adam also avoids the word insula when describing his Birca. And Björkö in Lake Mälar does not lie in an inlet of the Baltic directed towards the north, nor could the ancient harbour of Björkö be described as such an inlet or bay. Lake Mälar itself is certainly not directed towards the north as seen from the Baltic. Saltvik on Åland, on the other hand, is situated by an inlet (firth) of the Baltic that ranges to the north from its mouth. The information given in this passage is quite incompatible with the Birca on Björkö.

Item 3: (Book I, Chapter 60) "[The port of Birca] is highly esteemed by the savage peoples living around this sea, but it is rather dangerous for all those who are incautious and ignorant of such places." (... optabilem, sed valde periculosum incautis et ignaris eiusmodi locorum). "[The Bircanians were often afflicted by pirates] and they have therefore blocked this inlet of the troubled sea with hidden masses of rocks along more than 18 km" (Qui sinum maris impacati per centum et amplis stadia latentium molibus saxorum obstruentes ... fecerunt).

Comment: The statement about the difficult sea-approach to Birca sounds reliable enough. But it is unthinkable that the shallows causing the difficulties could have been man-made. Either Adam has rendered a deliberate lie, or he has drawn the wrong conclusion of what he has got from his informant. A strict interpretation is not acceptable. Still, we may note that all the sea-approaches to Saltvik lead through narrows that connect the Baltic with the large bay called Lumparn. These narrows are situated some 100 stadia from Saltvik. The approaches to the narrows are (and were) filled with shallows.

Item 4: (Book I, Chapter 60) "On this anchorage, being the best sheltered within the maritime region of Sueonia, all the ships belonging to Danes alias Northmen as well as to Slavs, Sembs and other Scithian peoples use to convene every year for sundry necessary commerce." (Ad quam stationem, quia tutissima est in maritimis Suevoniæ regionibus, solent omnes Danorum vel Nortmannorum itemque Sclavorum ac Semborum naves aliique Scithiæ populi pro diversis commerciorum necessitatibus sollempniter convenire).

Comment: A strict interpretaion of this passage is rather incompatible with the Björkö theory. This is so because Lake Mälar, although at the time connected and on a level with the Baltic, yet must have been filled with freshwater (at least on the surface). A Viking Age sailor, observing its seclusion and tasting the water, would hardly have regarded the Mälar estuary as a "maritime region".

Item 5: (Book III, Chapter 77 with skolion 94) The archbishop consecrated six bishops for Sweden, Adalvard and Acilin, also Adalvard and Tadiko and furthermore Simeon and the monk Johannes [alias Hiltin]. In skolion 94 is added: Adalvard Sr. was to superintend both lands of the Goths, Adalvard Jr. was ordained for Sigtuna and Upsala, Simeon for the Ski-Finns and Johannes for the islands of the Baltic. (... Iohannes ad insulas Baltici maris destinatus est). This information must be coupled together with an item in Book IV, Chapter 20 that reads: "For this city [i.e. Birca] he ordained, as the first among our people, the abbot Hiltin, whom he wanted to call Johannes." (In qua civitatem primum ex nostris ordinavit Hiltinum abbatem, quem ipse voluit appellari Iohannem).

Comment: The fact that a bishop was sent to Sigtuna makes it utterly improbable that anothear bishop should at the same time have been sent to a "capital" (civitas)7 on the island of Björkö, situated only 35 km from Sigtuna, that is called civitas maxima in Book IV, Chapter 25. It was unusual that two capitals were situated so near each other. Furthermore, the two statements about the destination of Johannes clearly indicate that the Birca referred to, would have been located on one of the islands in the Baltic. A strict interpretation of these passages is quite incompatible with the Björkö theory. (Besides, the town on Björkö was deserted and destroyed at the time of Johannes, see below.)

Item 6: (Book IV, Chapter 14) "Turning from the northern parts to the mouth of the Baltic we first meet the Northmen, then the Danish region of Skåne stands out, and beyond these live the Goths for a long stretch all the way to Birca." (... et supra eam tenso limite Gothi habitant usque ad Bircam).

Comment: The Goths of Sweden have never inhabited the area north of the bordering forest called Kolmården, about 90 km SW of Björkö. Obvioulsy, this information (and that of Item 7) must refer to some Birka orbirk (i.e. market town) other than the one with the missionary station. None of the missionaries appears in this passage, nor in the next..

Item 7: (Book IV, Chapter 23) "Beyond it [i.e. Western Gothia] Eastern Gothia extends along the sea, that is called the Baltic, all the way to Birca." (Deinde Ostrogothia protenditur iuxta mare illud ... usque ad Bircam).

Item 8: (Book IV, Chapters 16-17) "Further within, there are even more islands dominated by the Swedes. The largest of these is Churland. [...] We have been told that several other islands are in this sea, among them a large one called Æstland. [...] This island is said to lie near the Land of Women, while the before-mentioned lies not far from the Birca of the Swedes." (... cum illa superior non longe sit a Birca Sueonum).

Comment: What is called "Land of Women" (terra feminarum) is generally supposed to have been situated on the eastern coast of the Gulf of Bothnia, and consequently some 400 or 500 km from Æstland (i.e. Estonia). The distance between Churland (Latvia) and Björkö is about 300 km, (the same as between Latvia and Åland). We recall that the distance between Saltvik and Uppsala is 250 km by sea (and only 120 km as the crow flies), thus less than between Saltvik and Latvia. If one of these distances can be characterized by the words non longe, so can the other.

Item 9: (Book IV, Chapter 20) Adam first describes the coastal areas of the Baltic as inhabited by various nations. "In pity of their errors our archbishop ordained as their diocesan capital Birca, which is in the middle of Sueonia facing Jumne the capital of the Slavs and equally distant from all the coasts of the surrounding sea." (Quorum errori condolens noster metropolitanus statuit Bircam illis gentibus metropolem, quæ in medio Sueoniæ posita contra civitatem Sclavorum respicit Iumnem paribusque spaciis omnes illis ponti amplectitur horas).

Comment: The statement that Birca is equally distant from all the coasts of the Baltic is totally incompatible with the Björkö theory. Instead we may note that Åland lies roughly in the middle between the coasts of Finland and Sweden. A certain passage in the Flateyjarbók may be interpreted as describing the dominion of the Swedes in the early Middle agee as extending from the region north of Lake Väner in Sweden all the way to the region around Tampere in Finland (see below). Åland would thus have been somewhere in the middle of this dominion--at least as far the purely maritime region is concerned.

The city of Jumne is descibed in greater detail in another passage by Adam (Book II, Chapter 22). It was situated on the island of Wollin, in the mouth of the river Oder in Poland. The only place facing Jumne was the opposite shore of the arm of Oder called Dziwna, that separates Wollin from the mainland. None of the towns of the Swedes or the Goths, nor any place called Birca has been situated on this place, as far as is known. It is unthinkable that any source could have told Adam that Björkö lied opposite Jumne on Wollin. The market place Saltvik, on the other hand, was situated on the eastern shore of the long inlet previously mentioned. On the western shore of the same inlet there is a place called Jumala (at least since 1351), a name that may originate from an inflected form of the Estonian word juum or joom, This Jumala (now spelt Jomala) hardly belonged to the Slavs, but it may have been an Estonian village (which perhaps was all the same to a German bishop). Apparently Adam has caught some kind of misunderstanding here, and we cannot know for sure what it was. In any case, there is nothing in this passage that tallies with the Björkö theory.

Item 10: (Book IV, Chapter 20, skolion 126) "For those who sail from the Skåne of the Danes to Birca, the journey takes five days, from Birca to Russia likewise five days at sea." (A Sconia Danorum navigantibus ad Bircam quinque dierum iter est, a Birca in Ruzziam similiter per mare habes iter quinque dierum).

Comment: The distance from the nearest corner of Skåne to Björkö in Lake Mälar by way of the then still usable fairway at Södertälje was about 455 km. The distance from Björkö to the nearest corner of Russia, i.e. the mouth of Neva, is about 750 km. These numbers are related as 3 to 5 or as 5 to 8, not as 5 to 5. Item 10 is therefore incompatible with the Björkö theory. The distances from Saltvik to Skåne and Russia are, on the other hand, about 630 and 605 km, respectively. Suppose that Adam counted 124 km as one day's sailing. That would have meant 5.1 and 4.9 days for the two distances, which Adam's informant would certainly have abbreviated to 5--if the slight difference had been noticed at all.

Item 11: (Book IV, Chapter 20 with skolion 127) "For this city [i.e. Birca, named just before, see Item 9] he ordained, as the first of our people, the abbot Hiltin, whom he wanted to call Johannes" (In qua civitatem primum ex nostris ordinavit Hiltinum abbatem, quem ipse voluit appellari Iohannem). The skolion adds: "There is the port of Saint Ansgar and the tomb of the holy Archbishop Unni, and a familiar haven, it is said, for the holy confessors of our diocese." (Ubi est portus sancti Ansgarii et tumulus sancti Unni archiepiscopi, familiare, inquam, hospitium sanctorum nostræ sedis confessorum).

Comment: Here it is said quite clearly that Johannes was ordained to the very place where Ansgar landed and where the tomb of Unni from 936 was to be seen. The town on Björkö was at this time waste and deserted (see below) and could not have been the see of Johannes. Item 11 is totally incompatible with the Björkö theory.

Item 12: (Book IV, Chapter 29) "With the assistance of delegates from the famous King Stenkil [1061-66] he [the archbishop] established for him [Adalvard Jr.] a bishop's see at Sigtuna, which is a single day's journey [35 km] from Upsala" (Cui etiam per legatos clarissimi regis Steinkel sedem posuit in Sictonia civitate, quæ distat ab Ubsola itinere diei unius). "One gets there like this, from the Skåne of the Danes one reaches Sigtuna or Birca after five days at sea, for they are near [or alike]." (Est vero iter eiusmodi, ut a Sconia Danorum per mare velificans V° die pervenias usque Sictonam vel Bircam, iuxta enim sunt). "But by land from Skåne through the land of the [West-] Goths and their city Skara, Telge [Södertälje] and Birca, one reaches Sigtuna only after a full month." (Si vero per terram eas a Sconia per Gothorum populos et civitatem Scaranem, Telgas et Bircam, completo mense pervenies Sictonam).

Comment: The Birca mentioned in this chapter fits well with Björkö in Lake Mälar. The information could scarcely refer to any place on Åland. It should be noted, however, that the market town on Björkö lay waste since long before the reign of King Stenkil. We have also noticed that Bishop Johannes was ordained for a place called Birca at the same time as Adalvard was sent to Sigtuna (see Item 5). Without himself realizing the inconsistency, Adam obviously uses the name "Birca" for two different places in his Chapters 20 and 29. Naturally, both these places may have been called "Birca" (birk) in his time--especially if both were market towns.

Item 13: (Book IV, Chapter 30, skolion 142) "During his journey he [Adalvard Jr.] seized the opportunity to make a detour to Birca, which is now reduced to lone-liness so that one can hardly find vestiges of the city; therefore it was impossible to find the tomb of the holy Archbishop Unni." (Tunc etiam occasione itineris divertit Bircam, quæ nunc in solitudinem redacta est, ita ut vestigia civitatis vix apparent; quare nec tumulus sancti archiepiscopi Unni inveniri potuit).

Comment: This item obviously deals with the market town on Björkö (about 30 km from Sigtuna), which lay waste in the 1060'es. (The devastation has been verified by the dating of the archeological finds, see below). Adalvard must have believed that Unni had been buried in this Birca, a place probably familiar to the inhabitans of nearby Sigtuna.

Apparently he did not consider the possibility that Unni might have been stationed at the same place as his colleague Johannes, wherever that was. If Unni really had been buried at the Birca on Björkö in 936, and if the grave had thereafter been the destination for pilgrimages, the tomb would most probably have been extant and observable, even when the settlement was devastated.

The archeological indications of Christian mission on Björkö and Åland, respectively.

Björkö in Lake Mälar has been the subject of repeated and exceptionally exhaustive research. No Viking Time town or village in Sweden or Finland has been at all so well sifted out and documented as this one. The dating of the finds tells us that the settlement started about 750 and existed as a market town until it was unaccountably deserted about 970. Thereafter it lay waste. There are no finds indicating the residence or grave of a king or bishop. No remains of churches have been found. About 2,500 graves are identified, but none of them shows any sign of having belonged to archbishop Unni. Generally, they represent pagan burial customs.

Apart from the fact that the town on Björkö was an important commercial center with a harbor during a period of some 200 years, there is nothing among the finds to substantiate the accounts made by Rimbert and Adam.

The name "Saltvik" was originally applied to a part of the long inlet mentioned above, and the meaning of the name is 'Salt Bay'. The content of salt in the water at this spot is even lower than in the brackish water around Åland, so the reason for the name must be something else. Most probably the sheltered site and the position in the middle of the large inland sea consisting of the Baltic, the Gulf of Bothnia and the Gulf of Finland made Saltvik a natural center for trade exchange in the Viking Age. The northern Baltic countries exported hides and train oil in exchange for (chiefly) common salt from Slesvig and other parts of the southern coast of the Baltic. It would be natural to name such a port "Salt Bay". The international character of the port is emphasized by the place name Sollox applied to a place in the parish of Saltvik. This name seems to be derived from the two Estonian words sool 'salt' and laks 'bay'. Probably it is just an Estonian version of the Swedish name Saltvik.

The church of Saltvik near the ancient harbor dates from the 13th century. Close by is the former thing place, where we notice a flat rock that may have served as a dais at the sessions of the moot (thing). The surrounding area abounds in graves from the period of AD 500 to 1000. Within the parish of Saltvik 2,385 such graves have been identified, a fourth of all Viking graves on Åland. Since only very few of the graves have been excavated so far, they provide no proof of a possible Christian influence in this period of time. Anyway, it is known that the present church, built with stone walls, has had a wooden predecessor (see below).

Other churches from the 13th century, such as that of Sund (4.5 km SE of the Saltvik church) may also have had wooden predecessors, of course. It is impossible to ascertain the date of the first church built on any of these places, because the oldest foundations are likely to be hidden under the present buildings. Therefore, it cannot be excluded that the first churches were built in the 9th century, for instance by Gautbert/Simon and Herigarius.

Since long ago, there has stood a peculiar tombstone next to the church of Sund, not far from Saltvik. The memorial is made of limestone and shaped like a cross of St. George. Embroidered crosses of this shape have been noted on pallia and gloves worn by e.g. Archbishop Bezelius of Bremen who died in 10438. In 1950 it was discovered that there were carvings on the edges of the stone cross9. The carvings look like a number of runes along with two Latin crosses and one large cross supplied with 13 rays extending from the centre of the cross. The rays are distributed in groups of 2, 3, 4 and 4 (counted clockwise) in each of the four quadrants formed by the arms of the cross.

On a certain document pretending to be the pallium letter of Archbishop Unni from about 92010 there is also a St. George's cross with 13 symbols added in the quadrants. The symbols look like the letter S and are distributed in groups of 2, 3, 4 and 4 (counted clockwise) in each of the four quadrants. Some scholars believe that the document dates from the 11th century, thus representing what its maker thought would befit this particular archbishop.

Supposing that the carvings on the stone cross represent runes, these would correspond to the letters UINI I ILIS, but one should remember that the rune U stood for the sounds w and v as well, and the rune I for both e and i. Long as well as short sounds were normally represented by single runes. No satifactory interpretation of these rune words has been published so far. The word UINI may be read as Wenni, i.e. the form of the name Unni that occurs in the pallium letter mentioned above. Alternatively, we may read veni (Latin) 'I came' and illis 'to them'. (The middle I remains unexplained.) There is no other tombstone of this type anywhere on Åland. In Finland two similar crosses have been found and on Gotland there is a tradition about a wooden cross in memory of Archbishop Unni (supposed to have been the apostle of Gotland)11. It should be emphasized that the earth below the Sund cross did not contain the mortal remains of Unni, when the spot was excavated in 1950. Says one of the supporters of the Björkö theory: "The Sund cross must be regarded as a cross erected to the memory of Unni's visit on Åland [...] It substantiates an ancient tradition that Unni did visit Åland"12.

The territory populated by Swedes according to the Flateyjarbók

The Icelandic Flateyjarbók from the 14th century contains an enumeration of all the lands of the Swedes in Saint King Olav's time (i.e. about 1000)13. It is said that Svíaveldi (the dominion of Swedes) consists of the lands of the Western and Eastern Goths, Värmland, Markir (later called Dal), the islands of Öland and Gotland in addition to Svíthjod sjálfri (the nation of Swedes themselves). The latter is said to contain five parts, i.e. Södermanland, Västmanland with Fjädrundaland, Tiundaland, Attundaland and a fith part described as thá er hinn fimmti Sjáland ok that, er liggr hit eystra me? hafinu ('then is the fifth one the Sealand and what lies the eastern along the sea'--probably meaning along he eastern sea, i.e. the Baltic). The fifth part could hardly be anything else than the lake district in Southwest Finland with the adjacent coastal area around Turku.

There is also abundant archeological and linguistic evidence of the presence of Swedes in Finland during the Viking Age as demonstrated by the present writer14. We may also note the statement of Einhard (770-840) in Chapter 12 of his Vita Caroli Magni, that Northmen are living on the northern coast of an unnamed gulf (which obviously meant the Gulf of Finland). If we realize that Swedes may have inhabited the land on both sides of the Gulf of Bothnia, Adam's words "in the maritime region of the Swedes" (in maritimis Suevoniæ regionibus) acquire a well defined meaning, and Åland would have been regarded as placed in the middle of this region.

Summing-up

This scrutiny shows that Saltvik on Åland complies with a fair number of the criteria concerning the geographic position of the Birca missionary station that emerge from the sources.

1. Saltvik is situated in a maritime region on one of the islands in the Baltic.
2. While traveling by land one can reach Saltvik only by crossing the sea.
3. Saltvik is roughly in the middle between the coasts on each side.
4. The name "Saltvik" intimates a port highly esteemed by the salt traders, a port that would consequently have been well known in Hamburg.
5. The port of Saltvik is well sheltered.
6. The abundant shallows in all the sea-approaches to Saltvik would have been a danger for all who were incautious and ignorant of the waters.
7. The port of Saltvik is formed by an inlet from the Baltic directed to the north.
8. The distances from Saltvik to Russia and to Denmark were equal.
9. Åland is detached from Upsala to such a degree that it seems possible that it could have had a ruler (rex) of its own in the 9th century. The royal name Björn appears in several place names on Åland (just as in many other regions).
10. On the western shore of the Saltvik inlet there is a settlement called by a name that could be mistaken for the name of Jumne, the city of the Slavs on Wollin.
11. Saltvik on Åland is well suited as the see for an envoy to both Swedes and "other peoples in the northern regions" (e.g. Estonians and Livonians).
Saltvik, however, lacks two of the criteria of the missionary station in question.
1. There is no other source referring to Saltvik as a birk(a) or as being called Birka.
2. Saltvik is not near Upsala, as compared to e.g. Sigtuna and Björkö.

The market town on Björkö fulfills the following criteria requried by the sources:

1. It was a commercial town with a port.
2. It was situated in the middle of the region settled by Swedes.
3. The island on which it was placed has always been called Björkö (Bierkö. Biäerkö, etc.) which would be rendered as Birca in Latin.
4. It was rather near Upsala.

The market town on Björkö on the other hand lacks a number of criteria that would be requried according to the sources:

1. It was not situated in a maritime region but in the well-known Lake Mälar.
2. Crossing the sea was not required in order to reach it when traveling by land.
3. It was not situated in the middle between the coasts of the Baltic.
4. The port was not well sheltered by nature.
5. The sea-approaches to Björkö were not obstructed by shallows or rocks.
6. The port at Björkö was not formed by a bay or an inlet directed to the north.
7 The distances from Björkö to Russia and to Denmark are not equal, the former distance being 65 percent longer than the latter.
8. There was never anything like a town Jumne that could have been facing Björkö.

Let us also look at the archeological evidence. In this respect Saltvik stands out as rather well equipped.

1. According to Ringbom15 there have been wooden predecessors to the stone churches in Saltvik and Finström (10 km from Saltvik). These could of course have been built centuries before the extant ones and could thus correspond to the churches built by Gauzbert and Herigarius. Ringbom asserts that the general archeological evidence indicates an early Christianization whether or not it was brought about by the missionaries from Hamburg.
2. On a steep cliff within 2 km from Saltvik, there are remains of an ancient castle measuring 150 x 250 m, which would have looked like a fortified town when it existed. This castle could therefore correspond to the civitas or urbs mentioned by Rembert.
3. Beside the church of Sund a unique stone cross has been found with signs having certain relations to the signs on a document pretending to be Unni's pallium letter. The signs have never been assigned to any other person than Unni. The stone cross could well be either an original tomb/memorial of Unni or a copy of the original (although certainly removed from its primordial place).
4. According to the studies by Roeck Hansen16 the neighborhood of Saltvik has been inhabited continually since the 6th century. All the three known missionaries, Ansgar, Unni and Hiltin/Johannes could therefore have resided in Saltvik.

On the island of Björkö, the archeological finds abound and have been thoroughly studied, but nevertheless, the kind of evidence that we are looking for is lacking.

1. No trace of any ancient church or residence building has been found on the island, although nothing has ever been built on top of the remains of the abandoned town.
2. There is no trace of anything that could have been rightly called an urbs or civitas beside the remnants of the market town itself.
3. No tomb, not even a trace of anything that could remind of Unni, has been found on Björkö.
4. The archeological finds on Björkö demonstrate that the town did not exist during the entire period of the missionary activity directed to Birca. Hiltin/Johannes, who was ordained for Birca, could therefore not have resided on Björkö.

Discussion

Locating the Birca of Rembert and Adam on the island of Björkö is a tradition of long standing. It probably originated already with Adalvard Jr. who was looking for Unni's tomb on Björkö. The so called Prosaic Chronicle17 from about 1450 settled the question for centuries to come by stating that Birca was on the island of Björkö in Lake Mälar (Birka han laa vppa ena öö j mælar, som hether birköö). Eric Dahlberg passed on the tradition in his Suecia antiqua et hodierna of 1717.

However, already National Antiquarian Johan Hadorph (1630-93) found it necessary to start excavations on Björkö in order to repudiate critics who pointed out that it would be impossible to prove the identity of Birca with Björkö18. No conclusive proof was found, but the critics were reduced to silence anyway as time went on. Nowadays all historical authorities subscribe to the opinion that the Birca of Ansgar and Unni was on Björkö. Naturally it is admitted that Hiltin/Johannes could not have resided there, and therefore the statement of his ordination for Birca is explained away with the assumption that his bishopric was called Birca although the see was placed elsewhere.

The same kind of subterfuge is used throughout in order to sustain the notion from the Prosaic Chronicle that has already gained the status of an "historic fact". Thus Rembert may have happened to use the word maria when he really meant freti. He may also have happened to use the word vicus when he had seen an oppidum and the word urbs when he had seen a citadellum.

Adam may have been ill informed when writing of Lake Mälar as a bay of the the Baltic that draws to the north. Since hidden masses of rocks along more than 18 km could not have been man made as Adam says, he may have invented the whole thing. And when Adam says that Johannes was ordained to the insulas Baltici maris destinatus est it may be just a liberal choice of words to mean that he was ordained to the old bishopric of Birca that certainly included some islands. Adam's statement that Birca was equally distant from all the coasts of the surrounding sea could be just another proof of his being rather ill informed about the situation in the nordic region. The same may be surmised about the misrepresentation of the distances from Birca to Skåne and Russia.

If we allow for Rembert and Adam to have been enough indulgent toward facts, we may place their Birca in Lake Mälar or in almost any other place of our discretion. The written sources then become rather worthless as material for the historian. Tradition, intuition and authority take their place. The present article aims at another paradigm in so far as the aspect of probability is concerned. If a historical source reports a certain circumstance to equal A, we may evaluate its reliability as very low and conclude that it be just as likely that the real circumstance equalled something else (it could be B, C, D, etc, all the way to Z). The point is, that even in this situation the probability of A is far larger than that of any specific alternative (i.e. of B, or C etc.). Therefore, even when dealing with unreliable or doubtful sources, we should build up the theory of what really happened on the assumption that the particular statements may mean what what they say--until we can show that a certain statement cannot mean what it says.

Using this paradigm for our study of Rembert and Adam, we find that they do not place their Birca on Björkö in Lake Mälar. It has been shown above that there is at least one place that fits their descriptions better than Björkö--even it still does not fit perfectly. This place is the tiny village that was later known as Saltvik on Åland. It is true that Saltvik has been called just Saltvik (and Kvarnbo) and nothing else in all the written sources since 1322. This does not prevent it, however, from having borne another name before 1322 or from having been spoken of as the birk, (i.e. as the market town). Even Rome was often called just urbs, 'the city'. We should also remember that Iceland is not called Gardarsholm any more, and that the names Östra Aros and Nieuw Amsterdam long ago have been substituted by Uppsala and New York. There is also a number of place names from the early sources that have vanished completely after having left just a single trace in the records, e.g. Agnefit, Baalagaard, Helluland and Vinland. Therefore, we should expect as a possibility that also the name of Birca at an early juncture vanished from the place on which is was once applied by Rembert and Adam.

Since Rembert and Adam obviously did not have the market town on Björkö in view when mentioning Birca, it must have been some other place. That should be reason enough for discarding the identification of Birca with Björkö and to direct future discussions to other places. Saltvik on Åland is demonstrably among those to be considered.

Footnotes:
1) Sawyer, Birgit and Peter, Medieval Scandinavia, Minneapolis 1993, p. 101.
2) Ibid., p. 151.
3) Quellen des 9. und 11. Jahrhunderts zur Geshichte der hamburgischen Kirche und des Reiches. Darmstadt 1961.
4) Nationallencyclopedin, Höganäs 1990, vol. 2, p. 591.
5) Bjørnbo, Axel Anthon, "Adam af Bremens Nordenopfattelse" Aarbøger for nordisk Oldkyndighed 1909.Kjøbenhavn 1909, p. 138
6) Niermeyer, J.F., Mediae latinitatis lexicon minus. Leiden 1976.
7) The word civitas in Adam usually stands for the capital of some region: Odense of Fyn, Roskilde of Sealand, Lund of Skåne, etc. See Adam av Bremen, historien om Hamburgstiftet och dess biskopar. Stockholm 1984, p. 331.
8) Braun, J., Die litugische Gewandung. Freiburg 1907, pp. 374, 642. - Ärkebiskoparna från Bremen. Borås 1986, p. 46-51.
9) Ålands medeltidsurkunder, vol. 1, Helsingfors 1954, p. VIII; Det åländska folkets historia, I;1, Mariehamn 1983,p. 197.
10) Curschmann, Fritz, Die älteren Pabsturkunden des Erzbistums Hamburg. Hamburg und Leipzig 1909, p. 37.
11) Ringbom, Åsa, "Sundskorset ett medeltida minneskors?" Väster om Skiftet. Åbo akademi, Dept. of History,
Report Nr. 8. 1986, pp. 31-34, 39.
12) Ibid., p. 38-39.
13) Flateyjarbók. København 1944.
14) Nordling, Carl O., The creation of Finnish and the Finns. Lidingö 1995.
15) Ringbom, Åsa, "Dateringen av Ålands kyrkor" Historisk tidkrift för Finland 79. 1994, p. 459-93.
16) Roeck Hansen, Birgitta, Township and Territory. Stockholm 1991.
17) "Then Gamle Swenske Crönica." Scriptores rerum Svecicarum medii aevi. I. Upsaliae 1818, p. 242
18) Janson, Sverker, "Fornminnesbeståndet Björkö" Ansgars Birka. Stockholm 1965, p. 36.

Back to Scandinavian History