sml logo Dark Ages Re-Creation Company sml logo
+ Bibliographies
+ General
+ Interpretation
+ Pastimes
+ Reviews
+ Techniques
This translation was done by Michaela Gibbion (known in the SCA as Maeva Eriksdottir). Thank you on behalf of DARC and the re-enactor community.

Viking Age crafts in Ribe - A summary

by Mogens Bencard, Ribe with assistance from Kristina Ambrosiani, Lise Bender Jørgensen, Helge Brinch Madsen. Ingrid Nielsen and Ulf Näsman

Geography of Ribe

The settlement Ribe is located about 6km off the coast at the shore of the Ribe Au that connects Ribe with both the Baltic Sea [through the Kattegat] and the North Sea behind the Isles Fanö, Mandö and Römö. The area around Ribe belongs to the most northern part of the wadden sea reaching continuously from the firths of Elbe, Weser and Rhine protected by a long row of islands North of Ribe the western coast of Jutland changes its character. The wadden sea recedes and there are no more islands protecting the actual coast. Two good sized flood plains, the Ribe Au and the Koenigsau, run into the most northern part of the wadden sea. The firth of the Koenigsau is mostly shallow with numerous shoals whereas the Ribe Au continue a clearly defined path into the wadden sea.

Taking the geographical position Ribe's into consideration with the topic of this article it is obvious that Ribe is located at the most northern point of the natural trade lines of the Frisian and that the Ribe Au is the most northern trade route used to connect with the jutlandic back lands.

While Hedeby was Jutland's portal to the Baltic Sea Ribe was its equivalent to the North Sea. Whether there has been a trade route across Jutland from Ribe during the early viking age similar to the Hedeby-Hollingstedt-Route or not is debatable.

History – the written sources

The number of written sources available on the early history of Ribe are numerous and extremely informative especially considering that Ribe is a Scandinavian settlement. Here is a listing of the most important accounts:

Ribe is first mentioned around the year 860 — during the last years of Ansgar. According to Rimbert, Angar's biographer, the king presents the bishop with a piece of land and the permission to build a church.1 Furthermore he also receives permission for permanent residence of a clergyman. Following Birka and Hedeby, Ribe becomes Ansgar's third mission in Scandinavia.

In connection with the synod in Ingelheim in 948 three danish bishops are being mentioned : the bishop of Schleswig [Hedeby], Århus and Ribe.2 This is the first account of danish bishops known to us. Odinkar the Younger of Ribe, a bishop in the ii . century, may have even reigned all of Jutland from his bishopric as there is evidence that the bishopric of Arhus was abolished which was vacant in Slesvig [this doesn't make any sense – perhaps a misquote ? Need look up original text for this]3

At the end of the 11th century Adam von Bremen describes Ribe as a town surrounded by water flowing in from the ocean over which you can travel to Frisia, England and Germany.4 This source not only describes a town with trade connections along the coat of the North Sea and across but also indicates that Ribe is located exactly where it is known since the 12th century. It is safe to assume that the aforementioned water describes the uniting streams of the Ribe Au as no other location in this area meets those requirements It is therefore proper to conclude from the written sources
  1. That Ribe is an urban settlement comparable in importance to the archaeologically more renown settlements such as Birka and Hedeby.
  2. That Ribe was not only an established settlement during the viking age but continued as a town. Adam von Bremen refers to Ribe as civitas.

The archaeological sources

Archaeologically the trade routes and other connections between Ribe and surrounding areas can be proven back to the roman empire. Roman coins were excavated at numerous places and in great numbers in the extraordinarily rich living quarters of the field of Dankirke southwest of Ribe.

The sequence of layers of this living quarter spans from the pre-roman iron age to around the year 500 and contains a wide variety of finds [ceramics, glass, jewelry, iron, bronze, silver, gold] as well as a total of 73 coins of the age of the roman empire. The following coins of more recent eras were found as well : 3 merovingian coins minted by Madelinus between 689 and 716 in Dorestad, 2 English Sceattas from ca 725 and 8 Sceattas of probable Frisian origin of the late B. century.

There is the possibility that the living quarters at Dankirke were the predecessor for the final settlement of Ribe, just as Helgö was for Birka5

There are several viking age finds at Ribe and surrounding areas whereas numerous excavations in the medieval city center did not provide any clues of either the time of Ansgars or following centuries that would have proven that Ribe developed at the same location where it is today.

Only when the excavated area was expanded to include the other side of the Ribe Au in 1970 did the archaeologists gather positive evidence. A series of probings in this area of town revealed a concentration of cultural layers near S. Nikolaj Street that lead to two greater archaeological excavations in 1974-1975 with support of the danish archaeological society.

<Image 1>
Fig. 1. Detail of a city map of Ribe. Circles indicate excavation locations that only provided medieval finds. Triangles indicate excavations that also discovered Viking age finds.

A total of about 600 m2 have been searched. The cultural layers were between 2m and 3m strong, however, not even half of those originated from the Viking age. About 60.000 finds have been made of which about half are to be counted into the earliest time period. The oldest phases of the settlement have to be divided into 4 main layers. The bottom layer consists of soil with traces of ploughing with an ard. A grey layer of sand covers this ploughing layer [or might be part of it]. This layer also provided remains of one, maybe two wells as well as layer of waste consisting of iron scoria.

The whole area was completely covered by a waste layer during the second phase which consists mainly of unrotten dung mixed in with settlement waste that clearly points towards a settlement with trade and crafts characteristics. Items that have been pressed into the ploughing tracks of the first layer indicate that there hasn't been much time between phase 1 and 2. Phase 2 and 3 seem to share an equally close connection. This is evident in the stratigraphy as well as in the items found in layer 3 that share both character/style and time period as the items of layer 2.

The sequence of layers, however, shows a different composition. Alternating there are thinner layers with a concentration of charcoal, ashes and top soil oftentimes in connection with a fire pit or hearth and other relatively thick layers of sand or top soil mixed with sand.

In this text we will refer to those thin layers as “Activity horizons” and the separating layers “filling layers”. The term “ground level layer” has not been used in regard to the activity horizon layers as there are no signs of wall constructions or similar structures.

Two such activity horizons were identified as a bead maker's workshop and a bronze caster's workshop.

Incidentally only 2 sunken houses were found in the excavated area in connection with layer 3; they are located at edge of the layer. A double row of posts with interwoven brushwood was found but couldn't be classified as to its function due to its relatively small size. It could either be the wall of a house or perhaps a fence section. Layer 3 was disrupted and destroyed in this area due to construction works in the 19th century.

It is very difficult to interpret the various layers that make layer 3. The biggest obstacle lies in the many construction works this area has been experienced that disrupted and destroyed the older layers especially the rubbish pits of the 17th century. A larger connected area has not been excavated and examined. However, it became obvious that in this rural simple area ingenious and artistic crafts like the production of beads thrived.

The activity horizons were layered on top of each other and thus tend to create peaks with waste filling the spaces in between. None of those peaks have been excavated fully. It was only possible to spot check through several of them. In one of those samples containing the beadmaker’s workshop, several of those activity horizons were evident.

Phase 4 basically provides another waste layer on top of which there are finally the layers of the 13./14th century. It is difficult to understand this particular layer as it is very rich in objects and lacking in traces of actual settlement.

The following theory can be presented without claiming final interpretation: The first 3 of the 4 phases appear to show an increase in activity at the location. A waste layer with signs of trade and crafts was created over a former field. In direct follows evidence of the continued presence of craftsmen and their workshops at the same location. A possible explanation could be that the excavated area was located alongside a settlement center and a such give proof of a continued growth of this settlement. This settlement of which we found traces of ploughed fields might not have been nearby, however, the finds of waste and the aforementioned wells indicate it can't have been far away.6

<Image 2>
Fig. 2. Detail of a wall in profile of the courtyard showing layers and phases. The drawing is simplified. Organic material was found in the layers of phase 2. (Photo and drawing ASR).

Regarding phase 2 it appears that the settlement has been near the waste area. However, two points need to be taken into consideration
  1. We cannot assume that the settlement of which we have found evidence in the layers of phase 2 is an expansion of the settlement in phase 1.
  2. It is also not possible to determine whether the waste was created by a continued or a periodical settlement. The large quantities of unrotten dung appear to indicate a continued settlement. It definitely indicates that there have been barns from which the dung together with other waste was collected and removed. The great amount of dung also point towards a great number of animals, possibly greater than needed for the diet of the population of the settlement.
Assuming that at this early point in time Ribe did already have an established trade rout for Jutland cattle then this fact cannot be used to provide evidence for a continued settlement. This also applies to the amount of found dog excrements.

<Image 3> <Image 3b> 
Fig. 3 a-b. Detail of a profile wall in the backyard of the art museum showing layers and phases. The layers at the bottom of phase 3 include the activity horizons of the bead maker workshop (Photo and drawing ASR).

It appears that the work of the craftsman of phase 3 happened either under the open sky or a fairly lightweight construction as it did not leave any visible signs. This also appears to indicate a periodic settlement; however, this theory is weakened by the existence of fire pits obviously used for cooking and solid house constructions as living quarters.

Summarizing those finds it can be said that the excavated material is contradictory as well as too small in number to allow an educated decision as to whether Ribe was a permanent settlement or rather a periodically used market location. It is obvious; however, that it hosted a rather active economy.

Age determination

Due to the planned publication of those finds the variety of finds were divided into groups and handed over to a number of scientists for examination and study. Independently they have systematically examined the items in regards to the age determination which resulted in the discovery that they stem from the B. and g. century. The only exception is the waste of the comb maker workshop which originates mainly out of phase 2.

Kristina Ambrosiani, Stockholm, one of the scientists working on the project, reported a strong presence of combs with a D-shaped profile. Comparing this type of comb to similar material from Birka and Hedeby dates those most likely to the 10. century.7

Coins allow for a more exact age determination.8 30 coins have been found in the oldest layers : 28 Sceattas, 4 of which are so called "porcupines" and 24 of the "Wodan-Type" . There are two more coins of which one shows a hole drilled into the edge. Based on size they could be of Arabic origin, unfortunately they are heavily corroded which makes it impossible to determine this for sure.

<Image 4>
Fig. 4. The two main types of Sceattas. M. = 2:1. (Photo Nationalmuseum).

The coins are scattered throughout the layers and belong to phase 2 as well as phase 3 and 4. This arrangement also indicates that the coins are part of the settlement's waste rather than a scattered coin hoard. This is why they are valuable for the age determination. Kirsten Bendixen comes to the following conclusion : “After comparing treasure finds and scattered coin finds of the same time period it can be said that the different types of sceattas found in Ribe originated from Frisia where they were in use from ca 720 to 800.”

Since we were able to date a few coin finds from the lowlands to the middle of this time period we are confident that after consistent study of coin variants we are able to determine the age of the Ribe coins even more closely. The characteristics of the scattered coin finds of Ribe give the impression of a fairly widespread and common use of money, however, sceattas haven't been in use much longer after the coin reformation of Charlemagne in 790. Hedeby imitated the big thin Schroetling of the Carolingian coinage since 825 where it acquired some resemblance to the former sceatta. Actual sceattas have not been found together with Hedeby coins. This strengthens the theory that the use of sceattas was limited to a very small time period. It is therefore possible to date phase 2 and 3 to the time period between 720 and 825 with the phases dating more into the latter half of the 8th century and reaching into the following century. Those relatively exact methods of age determination – such as dendrochronology - of such a rich and diverse find such as Ribe can hopefully be beneficial to the study of other settlement excavations.

Immediately on top of the earliest layers of phase 4 are horizons, which point toward the 13th and 14th century. We are not going to discuss the problems that the lack of continuity of the finds throughout the viking age creates but would like to mention that scattered finds of the early viking age do occur in a different area of the settlement north of the Ribe Au. Also, at this time there is no answer to the question of where the hypothetical city center was located. However, west of the aforementioned excavation in 1976 a smaller excavation with scattered and sparse finds that were no different in content and date provided us with a completely different order of the cultural layers and horizons.

Comparing the written sources with the currently available archaeological sources we reason that the are of Ribe was continuously settled since the 8th century. A placement shift of the settlement as we know it from other earlier settlements in northern Europe seems very unlikely for Ribe, except for a shift during the 12th and 13th century from the northern towards the southern shore of the Au. With begin of the 14th century the oldest part of town is always referred to as »Suburbium«9

The various crafts

As mentioned before the various finds were divided between the scientists who agreed to publish both material and the results of their studies. For those find categories that are still under examination the following report is a preliminary account complemented by purely archaeological-stratigraphical observations by Mogens Bencard.

The blacksmith10

There is plenty material that provides proof for this particular craft from dross to fragments of burnt clay with bits of dross still clinging to it. Since this material hasn't been studied yet, this report is of a very cursory nature.

Aside from scattered finds the excavated material is concentrated at two points of the settlement of which one refers to the transitional time period of phase 1 to phase 2 and the other one to phase 3.

The waste appears to be the result of the workshop of just one smith and there is no evidence that indicates that there has been an extraction [working with ore] locally. The kind of dross mainly found is the flattened dome shaped “calotte” 11 dross puttied with roughly crumbled white rock

<Image 5>
Fig. 5. Front of a flue stone made from clay M = ca. 1:3. (Photo ASR).

Since such admixtures occur with a variety of dross it is difficult to discuss those found in Ribe as a stand alone kind. There is also a great percentage of another kind of light grey, porous and spongy dross which isn't documented for Hedeby and which cannot be connected to a single particular craft in Ribe. In Lindholm Høje near Aalborg this kind of dross was found in direct connection to the work of the local smith.

In his study of the material of Hedeby Robert Thomsen managed to select a type of dross which he referred to as “winkelförmige Esseschlacken” [literally “angle-shaped flue dross”]. The material of Ribe does not support this type of dross as its production is rather confusing. Among the finds of Ribe there are many examples of calotte shaped dross molten together at an angle with pieces of red clay with holes. In a number of cases the piece of clay was fully conserved and shows a rounded edge. The front of this item made of clay is glossy like glass, the back fired with a weathered texture. After noticing those attributes it was possible to put several of the fragments together which then create an approximately round shape with a hole in the center. It is difficult to see anything but some kind of protective shield for the bellows in those pieces, as stand alone flue stones. During use a partial burning of the clay occurs whereas the other part that doesn't get quite as hot flakes off more easily and after use rather weathers. This creates an archaeological item which is referred to as “flue dross” but is actually an independent flue stone – similar for example to the Snaptun-stone12 or the soapstone fragments from Hedeby and Fyrkat.13

One can imagine how those clay disks could have been a cover layer for the front of an actual flue stone as Robert Thomsen hinted at in his reconstruction. This theory is contradicted partially by the fact that the clay disks show a weathered back, that soapstone does hold up well against high temperatures and also partially the fact that the stone from Hedeby shows burnt in dross at the front. In case this interpretation proves correct it means that the angle shaped dross consists of two pieces: a calotte shaped piece of iron dross molten together with a piece of flue stone made from clay.

Bronze casting14

A great number of fragments of flue stones made from fired clay give evidence of this craft. To date there is no exact number of those fragments. There are also more than 300 fragments of crucibles as well as more than 2000 fragments of molds made from clay of which about 200 are decorated. 25% of this material were found in the layers of phase 2, 5% are scattered finds of phase 3 and the remaining 70% belong to the activity horizon of phase 3.

The activity horizon of phase 2 did not provide any firm evidence for the fact that the bronze caster did actually work at this place since for example no furnace had been found. However, in the center of the activity horizons there is a shallow depression with where convex dross covered with a flue stone has been found. A possible explanation is that the craftsmen did not need a more elaborate and bigger furnace than this. A small depression in the ground or maybe a mere piling of charcoal into which the crucible was placed with bellows and flue stone arranged besides it might not leave any clear archaeological evidence. The strong concentration of waste in this 2m x 2m area as well as a decrease of the same waste in the neighboring layers, however, makes it very likely that this has been the workshop of a bronze caster.

Aside from remains of bronze in crucibles only little bronze remains have been found in the area of the workshop. A connection between molds and bronze objects found on site could not be verified. There are many bronze items in the layers of phase 2 and 3. Only a very small number of bronze items have been found – a few needles, 3 triangular brooches, one small bowl-shaped brooch as well as fragments of a bigger one. The number of undetermined bronze fragments

<Image 6>
Fig. 6. Oberkappe (links) und Unterform für eine schalenförmige Spange. M = 2:3.

The great number of finds helps us understand the work process better.15 It not only applies to the smith but also to the bronze caster that there is one item that requires special attention. The many fragments of fired clay are not fragments of the furnace but flue stones that provided a protective shield for the bellows. In two cases it was possible to reconstruct and put together complete front sides of those items/ In Ribe those are predominantly square with an average side length of 5-6cm. There is a hole in the center of about 1cm diameter through which accommodated the tip of the bellow. Among the finds was a iron tube which could have been the tip of a bellow. The front side of the flue stones which pointed towards the fire are 1-2cm deep burnt to glass. The backside is not preserved as it received too little heat and thus crumbled away over the centuries. This type is known from Helgö, Birka and Paviken in Sweden and some burial finds from Sweden with examples up to 15cm strong. This item can also be understood as a clay block that sat on the ground.

There are 4 types of crucibles.
  1. Thimble-shaped crucible
  2. Thimble-shaped crucible with a square handle to the side
  3. Closed crucible
  4. Low and flat crucible16
There are only a very few examples of the type 4 crucible known and it is difficult to determine the percentage of the first two due to fragmentary nature of many. The crucibles of type 3 are the ones that are the least fragmented. Their heights range from 3.1 to 9.8 cm with an outer diameter of 2-5.7cm which indicates that they were only used to melt the bronze for one single item.

The many fragments enable the scientist to reconstruct the structure of a mold with greater accuracy. The finds do not allow any information on the actual model or the material out of which it was made but we will look at this later in greater detail. The molds were shaped from clay mixed with fine sand, glimmer and organic material [for example seeds and chaff] and fired. They are between 1-2 cm thick [average 1.5cm]. All fragments belong to two-piece molds except for a fragment of a three-piece mold.

The assembly went as follows: first the model was equipped with a conical peg that later formed the spigot for the finished mold. The clay of the top of the mold was modeled on to the top of the model, most of the time worked in layers, so the details of the ornamentation could be executed in greater accuracy. After drying the top was turned over and the model removed and incisions made into the area around the spigot. A piece of cloth approximately the thickness of the desired piece of jewelry was placed inside the top part of the mold and soft clay was pressed into the shallow depression to form the bottom. After drying the bottom part of the mold both halves would be separated, cloth and peg removed and if necessary holes for the needle clip cut into the bottom part of the mold. Before using the mold it had to be fired. This was done after putting both halves together and covering them with a thin layer of clay – a way to ensure they were fitted tightly together. We are not sure whether the molds were used several times. At least the bottom half of the molds had to be broken to remove the final piece. This is obvious in a small indention at the lower part.

The fragments of molds of phase 2 show that keys with a round handle were cast as well as other items such as needles. Only one single top half for a bowl-shaped brooch of the small type has been excavated. This type is also known from finds made on Bornholm. A brooch of this type has been found in Ribe in the same layer, however, it cannot be determined whether this brooch was manufactured in Ribe or not as no decoration on the inside was preserved. This may be caused by the structure consisting of thin layers which might have led to a flaking off of the inside layer.

The approximately 200 decorated fragments of the actual bronze casting workshop represent only a few different types. Among the finds is a mold for a horse buckle, a mold for an equal arm brooch of the Troms type as well asmoulds for two keys. The remaining decorated fragments belong to the type of bowl-shaped brooch of the Berdal type D.

The elements of the ornaments of those molds are the same; however, they are arranged in different ways so that the decoration changes from brooch to brooch. The bronze caster could have use the same top mold several times. It is however likely that he did not work with a constant model for example a finished cast brooch. The method with the greatest possibility is the use of an individually worked model made from wax.

The Berdal brooch is well known form from western Scandinavia. To date it has only been possible to tie one single brooch find from Myklebostad in Norway to Ribe which is likely to have been manufactured there. It is one of the mysteries of this time that there aren't many found items that match the many discovered molds.

At last we would like to point out that there have been many finds of iron pins for use with brooches. A stock of pins appears to have been a typical part of stocked items for a bronze casting workshop.

The production of glass beads17

Aside from clay glass is the most common material found during the excavations at Ribe. It appears in phase 2 as well as scattered finds during phase 3. More so, two activity horizons of phase 3 have been classified as workshops of bead makers.

<Image 7> <Image 8>
Fig. 7-8. The two furnaces of the bead maker workshops back-to-back. The lighter colored strips of the wall indicate a hearth whose function hasn't been determined yet. In front the activity horizons have been broken by a well dating from the Renaissance. (Photo ASR).

In total, ca. 1800 pieces of glass have been found. Of those 26% are whole and fragmented beads 20% glass rods, 21% tesserae, 20% waste and 13% shards of glass containers. In addition the following have been found : a piece of rose quartz, a piece of quartz, a piece of rock crystal, a bead of amethyst, a carnelian bead as well as 2 fragments of roman cameos with engravings of figurines18

The aforementioned activity horizons contain a strong concentration of glass. They are directly on top of each other separated only by a thin layer of sand. Both horizons have a length of 4,40 m whereas it hasn't been possible to determine the width as part so the area have been disturbed and removed through later activities and some are located outside the excavated field. Each horizon has a furnace in which the concentration of glass and charcoal are the highest. It measures 50-60cm x 25-30cm. Both are partially scorched red. In the less scorched parts of one furnace there was evidence that the clay had been mixed with refractory clay. Both furnaces showed an even surface without any depressions or elevated edges. In one corner of the one furnace a burned item made from fired clay was found that resembled a loom weight. It was slightly heavier than regular loom weights and covered in molten glass on one side. It is very likely this is a flue stone for the bead maker's furnace. (Fig. 7-8).

Of the ca 480 found beads 76% are single colored and 24% multi colored. 39% of the single colored beads are blue, 24% green and 17% white. The remaining 20% are red, orange, yellow, purple and clear glass; a few silver and gold foil beads have been found as well.

Fig. 9
Fig. 9. Diagram of glass items sorted by category. (drawing: Ingrid Nielsen).

The dominant color combination for multi colored beads is blue with decoration of red and white lines: 36%. There are another 24% of beads made from blue glass with different decoration. 14% were made from reticella rods, 12% from black glass with yellow lines and 5% millefiori beads. The remaining 9% are black, white, green and yellow beads with different ornamental decoration. Multicolored beads are predominantly round whereas single colored beads come in a much greater variety of shapes.

Generally speaking the beads represent the 8th and 9th century and west Scandinavia in style19. Only a few types of bead in very small numbers can be counted towards imported goods whereas the rest can be considered as produced in Ribe. Polished and drawn beads such as the gold and silver foiled beads are most likely imported goods.

Most of the single colored beads have been created in roll technique ie wounding the glass around a mandrel. Those beads could have been produced in Ribe. This also applies to the multi colored beads with either colored thread decoration as well as those from reticella rods created in the same technique. Millefiori beads are a sparsely represented group; however, the finds indicate that they have been produced in Ribe.

Of ca. 360 rods and stringers 59% are single colored and 41% multi colored . Of the single colored ones 39% are blue, 24% green, 17% red, 10% yellow, 9% white and 1% clear. Of all multi colored rods and stringers 44% are millefiori, 34% simple rods with parallel differently colored stripes and 21 % reticella.

Aside from the numerous finds of ready made single and multi colored rods and stringers there are also “collections of stringers” that show traces of manipulation with pliers. A “collection of stringers” refers to a number of multi colored stringers that are combined and molten into one thicker rod that can be stretched and thus thinned to the desired thickness. Both the existence of single colored as well as multi colored stringers indicates they were produced in Ribe rather than imported.

Of ca. 370 pieces of tesserae 21% are made of blue glass, 21% of red, 15% of green and 15% of clear green glass with encased gold foil. The remaining 20% are divided between clear, white, black, orange, brown, yellow and purple glass.

<Image 10>
Fig. 10. Steps of the production of millefiori beads. a: unfinished rod with blue threads and red and white flower pattern b: end of rod with traces of pliers. Blue and white eye patterns with red frame c: rod with eye patter in red-yellow-blue-white-blue. D: rod with blue and yellow checker board pattern in red frame e: plate with alternating blue and yellow checker board pattern and red and white eye pattern in blue frame. f: Bead with checker board pattern of aforementioned plate and blue and white eye pattern in red frame. Scale 1:2 sketch : P.-O. Bohlin).

The existence of tesserae is considered firm evidence of a bead production [or eventually the use of enameling for decoration]. The finds of Ribe indicate clearly that tesserae were a trading merchandise.20 Scattered individual finds for example a single piece in Aggersborg, are somewhat inexplicable, however, they could have been considered a curiosity or may have been used as an amulet. Wherever we find considerable amounts of it it must be considered an imported raw material for the production of glass items. So far there are no finds dating earlier than the 8th century. Already published finds fit neatly into the results of the Ribe excavations and draw a clear picture of the budding Norse trade due to efficient organization of trade routes with hubs such as Helgü, Paviken, Kaupang and the eldest of them, Ribe. A recently discovered find of 62 tesserae in the rural area of Stånga on Gotland can be considered evidence for the particular nature of gotlandic trade21

<Image 11>
Fig. 11. Examples of polychrome beads a: blue bead with red eyes and white zigzag line b: blue bead with red and white lines in a garland pattern c: blue bead of reticella and blue with red and white zigzag line d: blue bead with a blue eye framed in red and white. Scale: 3:2. (Sketch: P.-O. Bohlin).

In Ribe, tesserae appear to have been the most important raw material. It is possible that they have been produced for this express purpose. They were possibly introduced to Ribe via Western Europe but might also have been produced in Italy.22

The waste consists on one hand of remains of the imported raw material and waste of the bead production itself on the other. It is comprised of about 366 pieces of which 41% are blue, 14% multi colored, 15% green and 10% red with the remaining 20% of clear, white, black, orange, brown, yellow and purple glass.

The raw material was possibly imported in bigger blocks and broken into smaller pieces at the workshop before it was molten in crucibles. Numerous slightly bigger pieces of the same glass of which tesserae were made were among the finds. Therefore you find many broken pieces of tesserae between the smaller shards of glass as they were being prepared for the melting process.

The production waste consists of molten drops of glass encrusted with sand as they fell onto the sandy floor of the workshop still soft and thus sticky. Among those drops of molten glass some have been found of clear glass with remains of gold hinting at the fact that also tesserae with gold foil have been molten and reshaped. Shards of glass for example a reticella shard are also among those re-molten and re-shaped drops.

<Image 12>
Fig. 12. Examples of half finished or failed beads a: blue bead with center hole still filled with clay separator, one end partially molten b-c: a blue and two red beads with traces of being rolled around a metal spindle d: green bead made of a wound rod e-f: yellow beads made from tubular rods; two have molten together when the bead maker attempted to round the edges. Scale: 3:2. (Sketch: P.-O. Bohlin)

<Image 13>
Fig. 13. Examples of reticella rods and beads. Of the two rods a is made of blue glass with red and white veins (unfinished) and b is blue glass with red and yellow veins. The beads color scheme is blue, white and red. Scale: 3:2. (Sketch: P.-O. Bohlin)

<Image 14>
Fig.14. Examples of waste from production a-c: ends of stringers of green glass with traces of pliers d: millefiori of blue glass with parallel of blue glass with parallel white and red veins. e: 2 collections of stringers for the creation of millefiori rods made from red, yellow and blue glass. Scale: 3:2. (Sketch: P.-O. Bohlin)

Only 169 pieces of glass shards have been found. They are predominantly wall shards of glass items, 25% are shards from edges and 2% shards from the bottom of a glass item. In 82% of the shards the glass is light green, 11% dark green and the remaining shards are distributed between the colors blue, clear, olive colored, brownish red and purple.

Reticella and funnel glasses are the most dominant types of types of glass containers which means if is mostly drinking vessels that can be dated into the same time period. It is possible that there are also older types existent, however, if so then they only occur in insignificant numbers.

<Image 15>
Fig. 15. Shallow iron pan with even bottom and short handle that has been found in one of the furnaces of the bead maker. Scale: 4:5. (Sketch: Moria Mackenzie).

There are only a few hints that the bead maker used shards to create beads. The existence of glass shards is best explained as sign of extensive trade with glass drinking vessels that occasionally broke into pieces during their transport. Similar finds in Helgö and Birka have been interpreted in the same way.

We are not going to elaborate on the technique of bead production other than some discoveries made in regards to the tools of the bead maker. In the workshop layer we found a fist sized stone with a depression in the center which could have been a mortar. A shallow metal pan with an almost completely even bottom and a short handle has also been found in the hearth. It was possibly used to melt glass frit. A bone spatula was discovered in the workshop layer. An iron stick with conical tip and wooden handle is considered a scatter find; however, it might be possible that this is indeed a mandrel or other kind of tool from the bead maker's workshop. The traces of pliers on monochrome as well as reticella and millefiori rods have to be mentioned in regards to discussing tools.

<Image 16>
Fig. 16. Iron rod with conical tip and remains of a wooden handle. Length 29.5 cm. Was not found in context with function but is likely a tool of the bead maker. (Photo ASR).

To sum up, the following can be said about the glass trade:

It is the fact that in relation to the number of beads and shards of other glass items a much larger number of tesserae have been found that is characteristic for Ribe, in comparison with other places that show signs of bead production. The following shows what importance the proof of workshops have on the classification of a finds: The excavation was executed in two phases which each led to the discovery of about 50% of the total of found glass material. However, the last phase in which also the workshops have been discovered provided us with 76% of the glass rods. This find enables us to more clearly define and classify the find in its entirety.

The workshops were obviously of a very simple nature. Nothing indicates a solid construction with walls and roof. The floor was covered with a layer of sand and measured 4.4m. Unfortunately this is the only measurement available. The small furnaces of fired clay are on the ground. The bead maker possibly worked using a low table where he kept tools and raw materials. Remains of such a table have not been found.

Using crucibles he melted the glass in open furnaces. In order to do so a temperature of only 1000°C is necessary this can be achieved by means of bellows. A helper to work the bellows was necessary to produce a constantly hot fire. This helper might have been helpful during other steps of the work process as well; however, the main body of the work was done by the bead maker alone even though two craftsmen per furnace might have been the ideal cast. With the furnace immediately on the ground the bead maker must have worked crouched low and squatting.

Here follows an attempt at explaining the existence of several activity horizons on top of each other and separated by layers of clean sand. It is difficult to determine whether the individual layers should be considered as individual horizons for seasonal reoccurring workshops or as regularly happening floor reconstructions of a workshop with continued production. Since there are no signs of a house, the production possibly took place during the warmer season of the year, probably during the time when traveling merchants temporarily settled at a market place. However, there are also arguments against this theory since the excavated area might have hosted the work shop during the summer and that the actual workshop of the bead maker has yet to be found.

The large number of finds scattered all over the area of the excavations indicate that the two areas classified as bead maker work shops might not have been the only ones. The beads that can't be dated due to their type need to be examined as chronologically congruent with the other finds.

With the provision that this is only a preliminary report it appears that there has been ample production of glass beads in Ribe before and after 800 and that there was at least one workshop was active.

Up until today the production of millefiori and reticella beads has not been thought possible under as simple conditions as in Ribe. The discoveries of Ribe thus ends a traditional theory that for the complicated trade of glass bead making you had to look elsewhere like Italy or even Egypt and Syria. This also meant that this trade had been preserved in good sized workshops with a long tradition. The glass finds of Ribe however show that even more complicated and complex techniques are possible under the fairly primitive and simple conditions of a market place of pre-urban character.

With the aforementioned caveat in mind one can also assume that the bead maker was a traveling craftsman, possibly from Western Europe, who set up shop in Ribe during the trade season. This stay appears to have been of extended nature and created also a decent production based on the complicated techniques used in bead production.

It is unthinkable that the bead production was only for the local market of Ribe; however, it is near impossible to judge the number of possible customers from the back lands due to the lack of finds of Danish burial and settlement sites. Thorough studies on the prevalence of the various types of beads in regards to time and place are still necessary. As long as it is not possible to study other work shops the results of excavations will be very vague. In the list of craftsmen of this study in the northern world the bead maker from Ribe holds a very high rank.


A large amount [about 2.2kg] of untreated amber has been found. Among the amber finds were also 3 pieces that showed preparatory work to dome-shaped gaming pieces as well as several half finished beads and pieces with drilling holes. Amber was found scattered through the layers of phase 2 and phase 3. In connection to an activity horizon in phase 3 a sand island mixed with chips of amber was discovered. It cannot be determined accurately whether this was a workplace for a craftsman working with Amber who would have considered those chips as usable raw material instead of waste.


Only half finished items and very few finished objects made from wood have been found in Ribe and thus there is not much further reason for continued study from a craft perspective. Remains of wooden constructions and raw wood only show traces of fairly primitive processing with a hatchet. The bottom of a wooden well construction is put together by means of mortise joints. The mortises have been shaped with a hatchet and the hole at each end has been drilled with a spoon drill and shaped with a hatchet. This kind of drill is listed with items with iron components, but not the hatchets.

Four staves have been discovered in the same well. They were made from oak wood which despite its slow growth displayed 170-180 annual growth rings on its fairly small width. There has been an attempt at determining the age by means of a dendrochronological study, however, the pattern of annual growth rings could not be matched with that of the area of Slesvig-Jutland. While therefore the trade of coopers does not appear to be native to Ribe, that of the ship builder is a very likely possibility. Only one big iron anchor has been discovered in recent excavations but the many found tacks show the same kind of breakage as the finds Paviken on Gotland brought to our attention by Per Lundström.23

The antler, bone and horn trade24

The finds consist of about 1100 pieces as well as 5-6 liter of stag antler waste. It was mainly found in the unrotten dung layers of phase 2 which probably relates to the preservation attributes of this layer. The main item of the finds is production waste.

Only 35 pieces of actual items have been uncovered as there are 12 fragments of combs, 8 gaming pieces, 2 die, 8 Tibia with a hole drilled into them, 3 needles as well as 2 ”Schlittschuhe/ice skates”. Individual pieces can originate from a later period. The distribution of waste and finished items is characteristic for Ribe except for the bead production and not very surprising. In the excavated area items were produced, not used.


One of the pieces of this category stems from a moose; the remaining pieces are all from stags. Of the preserved antlers [all stag] about ¼ are still with the base whereas the remaining pieces come from antlers that had been shed naturally. This distribution resembles the distribution of finds from Hedeby and indicates there as well as here that not only antler from slaughtered animals were used but that they also went actively out to collect the raw material for the intended production. The characteristics of the antler waste in Ribe resemble that of similar settlements of the same time in the Baltic area. This indicates a similar work technique and therefore possibly similar products.

Dr. Ingrid Ulbricht in Sleswig represents the theory that there wasn't enough demand to justify a professional production during the Viking age and regards the production of combs as a secondary trade to a craftsman with a different main job.

However, the striking similarity of waste from place to place appears to point into a different direction. In Ribe a high concentration of waste, especially the shavings caused by working the raw material, was found in a clearly defined area. Despite the fact that no definitive work shop area was discovered the amount of waste still indicates a regular trade.

Apart from indirect evidence for the tools in use [knife, saw, polishing stone] provided by discovered items, only one single tools has been excavated that might have been used in regards to working with bone and antler. It is a square block of antler measuring 4.5cm wide and 1.5cm tall which provides the base to an iron point with a square profile inserted into its center. This item might have been used as a clamp of sorts to hold smaller objects in place while they were being shaped into dome-shaped gaming pieces for example. In finds from Birka matching indentures in the bottom of such pieces have been found.


The aforementioned pierced Tibia, needles and "ice skates" do not indicate that bone has been used in extensive professional production. This also applies to half finished pieces for the construction of combs cut from rib bones. However, there might already have been a trade based production of bone combs in place.


Items or waste from horn (Keratin) have not been found, except for pegs from horn of cow, sheep and goat. Those pegs do not show any marks of manipulation and thus aren't interesting in regards to this report on crafts in Ribe of the Viking age.

6 joined horn pegs, 66 single pieces and 12 smaller fragments of horn pegs from goat have been found. Considering the fact that only 6 bones of goat have been found one cannot doubt the theory that they have been collected specifically.

Those horn pegs do show marks of manipulation where they have been sawed or hacked off. This leads to the assumption that goat horns (keratin) have been used for the production of so far unknown items. Those horn pegs have been found in the same area as the antler waste.

Working leather25

Leather has been found exclusively in the layers of unrotten dung of phase 2 where it has been preserved. A total of 430 pieces of leather have been excavated. An attempt at determining the kind of leather based on texture and orientation of the hair follicle has been made. The majority of finds were cow leather, the remaining pieces from horse, sheep and goat.

<Image 17>
Fig. 17. 2 types of shoe. On the left "skin" shoe made from one piece and on the right, vamp for shoe with separate sole. (Photo: Marianne Carlsson).

Apart from the cut off tip of a knife sheath and a peculiar item resembling a leather key, the most remarkable leather finds are shoes of which 2 main types have been discovered:
  1. "skin" shoe made out of one piece of leather
  2. sole shoes where top and sole have been sewn together
An entirely preserved skin shoe shows that a 5mm strong leather strap has been used to sew it together. To either side of the foot opening there is a row of slanted holes used to pull a leather strap through to tie the shoe tightly to the foot.

Actual yarn has been used to sew the sole shoes together. Remains of the thread have been preserved and indicate it's been of botanical nature. The thread was pulled through holes that had been punched into the leather first. This kind of seam has a very characteristic appearance and can even be determined on fragments. Two complete uppers have been excavated, one the upper to a boot, and fragments with part of a seam. The uppers are cut from one piece of leather in a v-shape and sewn together at the heel. The technique is similar to the one in use during the later Middle Ages; however, without using the variety of sewing techniques with regards to the task at hand as it was the case in the later Middle Ages. All uppers show traces of the same decoration: a sewn welt from instep down to the toes. The shape of one sole indicates a pointed heel whereas two others have a rounded heel. This is also the case with the preserved boot and it is possible that shoes had soles that were shaped with a pointed heel.

The technique of cutting the pattern pieces is also different from the methods used during the later middle ages. While in the later middle ages the uppers were oftentimes pieced together, a method which guarantees the best use of the leather, both types of shoes found in Ribe were cut in a way that would create a rather great amount of waste.

The remaining leather finds are pieces of various sizes with marks of manipulation that indicate that they've been cut off along the edges. This marks them as production waste which in turn indicates that the leather work shops must have been nearby whereas the other items appear to be actual waste of the settlement. The waste allows no theory in what else was produced. A few triangular pieces of leather are probably related to the v-shaped uppers of shoes.

The sole shoes have been made mostly from goat and sheep leather whereas the skin shoe was cut from cow leather. The same kinds of leather dominate the waste. Especially the sole shoes appear to have been made by a professional shoemaker who managed to achieve a professional uniformity in his products. The waste indicates that there was leather trade in Ribe. It is therefore also believable that the excavated shoes were made in Ribe.


In regards to woven textiles a piece of roughly woven piece of wool and a finely woven piece of a vegetable fiber, possibly flax, have been found. Aside from this 100 loom weights and many fragments of fired clay have been unearthed. Furthermore weights of unfired clay were discovered but only a few exceptions of those were in a state that allowed complete excavation. 30 spindles of fired clay and one wooden spindle were found as well. These finds are distributed over the layers of phase 2 and 3. There is one area with a certain concentration of loom weights which were found scattered and not in one row. All other finds must be regarded as scatter finds.

The question whether there was a weaving trade in Ribe or simply production for their own use cannot be answered at this point.

One of the major points of discussion is about “how many” spindles and weights you need to find before you can consider them “many”. Considering the fact that for a loom you need at least 15-20 weights a total of 100 found weights is a rather unimpressive number. On top of that there are also the fragments and of course that fact that these finds are classified as scatter finds.

It is near impossible to compare this find to that of other settlements. Size of the settlement and the style of archaeological approach as well as time period of the settlements play an important role. However, while there's hardly any material for comparison there is the general impression that the concentration of this particular kind of find in Ribe is unusually numerous.

The zoological side of the studies if this excavation offers information that might be of interest in this regard. In comparison to other studies the bone material of Ribe has a greater concentration of bones from old sheep.27 This could indicate a greater interest in the production of wool than keeping sheep for meat. Since there appears to be evidence for sheep farming aiming for production of wool and therefore catering to spinning and weaving it raises the question whether the lengthy production process of making cloth was feasible for settlements of a temporary/seasonal nature, no matter whether the production aimed at covering the needs on site or as a trade.

A formula based on the number of found spindles and loom weights that would allow us to determine the kind of production level of a settlement would be very helpful to answer this question.


To sum it up, it can be said that the craft-related finds of the excavations clearly indicate that in the oldest part of Ribe almost all crafts known from the Viking age were active. Under this aspect Ribe can be classified as a fully developed settlement. The discoveries of the work shops for bronze casting and bead making allow significant and new contributions to the understanding of the development, the methods of production and the standing in society of those two trades.

Notable in regards to the bronze casting workshop is the fact that so much material only provided information on so few different types. This concentration of one particular kind of brooch in one work shop appears to indicate a specialization; a thought that'd open up a whole new horizon of theories. However, we cannot tell for certain whether the same craftsman tried for diversity in the production of other object categories or if he managed to hold on don to his specialization for his whole life. Comparing the limitations of the material of Ribe's bronze work shop with the great variety of shapes and types of bronze items of the Viking age it is surprising that not much more waste of bronze work shops has been discovered.

<Image 18>
Fig. 18. The weights for the loom are mostly plain. Extant ornaments are circled crosses and indented marks of keys of that type. Scale 2:3. (Drawing:.Hette Madsen)

While the glass finds haven't been studied under this aspect, the excavated material suggested a certain level of specialization. An example is the reticella rods and beads made form these as they are limited to this one work shop.

The special attributes of the layers of unrotten dung of phase 2 contributed greatly to excavating unusual organic material. Important are for example the evidence that horn matter [keratin] from goats had been used for the production of objects. The leather finds contribute to a better understanding of the craft during this time period since it is very likely that the shoes were crafted in a highly professionalized and specialized manufacturing process. It is astounding to find the primitive skin shoe in the same area as the fully developed shoe with separate sole and leather upper. This type hasn't been able to date as early as it has been the case here.

The conserved shoes undoubtedly were considered trash since they are completely worn out. It is typical for Ribe that excavation produce mainly waste but no finished items. An exception is the beads that once lost were hard to recover by the bead maker.

A major part of the production for example bronze items and beads are luxury items. It is therefore most likely that the finished items were buried with their owners. This means it is very unlikely to find them inside the settlement. The odd ratio of waste to finished items is very apparent in regards to the material found with the comb maker workshop. Combs are very common objects of every day use, although the “status” of the comb during that time period is being discussed, and one would have expected a higher number of used finished items during the excavation.

In general, the finds of the excavations indicate that the production of items of society of Ribe was not aimed at satisfying the local demand only. This discovery leads to the question about the overall character of the settlement. This problematic has been touched briefly in earlier chapters without trying to present evidence in a way that would not lead to the forming of a certain opinion on this matter too early.

The most striking attribute is the lack of evidence of house constructions and aside from the remains of the two sheds the overall lack of marks of settlement. This lack contradicts the richness and variety of excavated material of this area. Is this evidence enough to consider a periodical settlement with market character or was there a constant settlement after all?

Basically it is misleading to present this problem as an either-or question. The represented crafts cannot be considered indisputable evidence. They rather tend towards indicating a temporary settlement.

It is also possible that this kind of market place came into existence following a constant settlement. The question is thus best phrased like this: Does this find that does not show marks of settlement provide any evidence for a constant settlement ? With reservation the ploughing marks of phase 1 can be considered evidence for this idea. The aforementioned wells and the waste of the smithy do not appear in direct stratigraphical relation with the ploughing layer. While it is likely it is not certain that the ploughed fields and the settlement have been in immediate vicinity of each other. The recovered dung of domestic animals of phase 2 is another sign, even better than the ones mentioned so far. However, aside from the fact that the amount of dung can only originate from domestic animals in barns this is also not an indisputable evidence to support the theory of a permanent settlement. One craft, namely the production of cloth, appears to offer the evidence with the greatest importance. It is the most popular craft in settlements with agriculture and self-sufficiency and thus provides a good clue for the existence of a permanent settlement.

As mentioned before there is no probability calculus to determine the necessary amount of preserved tools before they are considered the remains of a true craft production with intended sales. In this regard the finds of Ribe share a striking characteristic that sets it apart from other prehistoric finds. The bone material shows an excess of older sheep, ie sheep were kept for wool and less for meat. This also means that the whole process from sheep to shawl is present. This process is lengthy and time-consuming and requires a certain settledness which contradicts the theory of a periodical market place.

In other words, it appears that the production of textiles though difficult to define as a professional trade offers the best evidence for a permanent settlement at the northern shore of the Ribe Au in 800 and before that.

(1) Vita S. Anscharii. Script. Rer. Dan. I, 482. 1772. (back)
(2) Dipl. Dan.1, I, 319. 1975. RICHERUS: Historiarum libri IIII. Recogn. Georgius Waitz, 73-74. 1877. Monumenta Germaniae historica, Scriptorum, III, 395-96. ADAM VON BREMEN: Hamburgische Kirchengeschichte (ed.B.Schmeidler), 64-65. 1917. (back)
(3) ADAM VON BREMEN: Op.Cit., 230-31. ANDREAS NISSEN: Danske Bisperækker, 56-57. 1935. (back)
(4) ADAM VON BREMEN: Op. Cit., 228-29. (back)
(5) ELISE THORVILDSEN: Dankirke und KIRSTEN BENDIXEN: Menterne fra Dankirke, Nationalmuseets Arbejdsmark, 47 f. 1972. (back)
(6) OLAF OLSEN: Nogle tanker i anledning af Ribes uventede høje alder, Fra Ribe Amt XIX, 240. 1975. (back)
(7) KRISTINA DANIELSSON: Bearbetat ben och benhorn, Birka, 40 f. 1973. (back)
(8) Die Münzen werden von Museumsinspektør Kirsten Bcndixen, Den kgl. Mønt- og Medaillesamling, Nationalmuseet, bearbeitet. (back)
(9) MOGENS BENCARD: Ribes ældste udvikling, Mark og Montre. 1974. MOGENS BENCARD und STINE WIELL: Et trefliget spænde fra Ribe, Mark og Montre, 1975. (back)
(10) Die metalltechnischen Untersuchungen werden von Konservator Helge Brinch Madsen, Konscrvatorskolen, København, durchgeführt. (back)
(11) ROBERT THOMSEN u.a.: Untersuchungen tiur Technologie des Eisens. Ausgrabungen in Haithabu, Bericht 5. 1971. (back)
(12) P. V. GLOB: Avlssten, KUML. 1959. (back)
(13) ELSE ROESDAHL: Fyrkat II, 45. 1977. Essesteine aus Ton kommen auch in Fyrkat vor, a.a.O.61. Ein schwedisches Beispiel gibt INGA SERNING: Förhislorisk järnhantering i Dalarna. Järnkontorets Forskning. Serie H, Nr. 9. 1973, 113 f. (back)
(14) Wie Anm. 10. (back)
(15) HELGE BRINCH MADSEN: Specialist i Spænder, SKALK, Nr. 4. 1976. (back)
(16) Typ 4 ist identisch mit den von Fyrkat erwähnten »heating trays«, siehe RoESDAHI.: Fyrkal II, S. 51. (back)
(17) Das \laterial wird von al. kand. Ulf Näsman, Viborg, bearbeitet. Dcm Kunsthandwerker Finn Lynggaard ist air Ratsrhlüge über herstellungstechnische Fragen zu danken. (back)
(18) BIRGITTE: \\'ISTOE`r: I'iklaria fra Ribe. Alark og Alonlre, 1978. (back)
(19) JOHAN CALLMER: Trade beads and head trade in Scandinavia ca. 800-1000 A.D. Lund 1977. (back)
(20) Vergleiche AGNETA LUNDSTRÖM: Bead making in Scandinavia in the Early Middle Ages, Antikvarisk Arkit 61. 1976. (back)
(21) DAN CARLSSON: Ett vendel - vikingatida verkstadshus på Gotland. Fornvãnnen. 1976/3-4. 86 f. (back)
(22) AGNETA LUNDSTROM.a.a.O. (back)
(23) PER LUNDSTRÖM: Kinknaglarnas vittnesbörd, Sjöhistorisk årsbok. 1971-72. (back)
(24) Das Material wird von fib kand. Kristina Ambrosiani, Stockholm, bearbeitet. (back)
(25) Das Material wird von cand. mag. Ingrid Nielsen, Sydjysk Universitetscenter, bearbeitet. (back)
(26) Das Material wird von mag. art. Lise Bender Jorgensen, Langelands Museum, bearbeitet. (back)
(27) Das Knochenmaterial wird von Museumsinspektor Tove Flatting, Zoologisk Museum, bearbeitet. (back)
      Updated: 4 Dec, 2007
Text © Translation Michaela Gibbion, 2006
Copyright details
Contact us