Høj Stene – a monumental ship setting by the Gudenå river

Høj SteneA monumental ship setting by the Gudenå river Ship settings, or stone ships, in the sense of stones arranged in a regular pointed-oval shape, are known especially from Denmark and Sweden, where some can be dated to the Late Bronze Age, others to the Late Iron Age and Viking Age. Of the latter, so-called monumental ship settings stand out by virtue of having a length in excess of 40 m (fig. 1). A further general feature is that, either singly or in groups, they occupy prominent positions in the landscape, usually by a main traffic route and often in association with barrows; some are found in conjunction with a royal residence – Jelling and Lejre. Insofar as rune stones form an integral part of a ship setting, as is said to be the case at Glavendrup and Bække and is suggested by the circumstances at Jelling, the monument can be dated to the Viking Age. In other cases, the dating of the monuments is less certain, with several possibly deriving from the end of the Late Iron Age. There is also great uncertainty about the function of these monumental ship settings. There has been a general tendency to perceive them as burial structures on a par with the other ship settings. But the preservation conditions at the archaeologically investigated monuments have been poor, and it has therefore not been possible to demonstrate a clear link between monumental ship settings and possible graves. Even so, these monuments can clearly be perceived as a power manifestation by an aristocratic environment, and they can be viewed in conjunction with the ship symbolism that is evident in Late Iron Age boat graves and the spectacular ship burials of the Viking Age. With the aim of addressing these issues, Moesgaard Museum has investigated the site of one of southern Scandinavia’s largest ship settings, located at Vejerslev, close to Gudenå river (fig. 1, no. 1). The monument was demolished more than 150 years ago. The first mention of it in written sources is in 1683 under the name “Højs Steen”, located on the western fields of the village of Vejerslev, which run down towards the Gudenå. A written source from 1768 states that the ship setting extends between two small barrows standing about 100 paces apart. A report submitted to the National Museum of Denmark around 1850 contains valuable information about the ship setting as the detailed text is accompanied by a survey of the ship setting’s ground plan. It is stated that the monument was c. 88 m long and 13 m wide in the middle. The drawing shows 16 preserved stones. These stood 2-4 m apart and were c. 1.9-2.5 m high (fig. 4). But the stone at the northern stem/stern, which was c. 4.5 m in length, lay toppled on the northern barrow. The distance from the Gudenå river is given as just less than 100 m. A written source from 1877 states that the final remaining stones were removed in 1852. In 2014 and 2016, systematic geophysical surveys were undertaken of the area between the mound that could be the northern barrow and the remains of the presumed southern barrow. Collectively, the geophysical surveys sketched a picture of an extended pointed-oval formation between the two barrows (fig. 6). The structure had been c. 82 m long and 14 m wide at its broadest point. This concurs well with the dimensions given in the report to the National Museum around 1850. The ship setting was oriented NNW-SSE and lay almost parallel with the Gudenå, c. 130 m distant from it. It was located on the edge of a level sandy plain that rises above the riverbed. An archaeological investigation was undertaken in 2016, with an excavation trench measuring c. 18 x 9 m being positioned across the southern part of the ship setting, in a place where the geophysical surveys had demonstrated stone traces or possibly stones (fig. 6). Of the four anomalies evident on the geophysical surveys that fitted with an arrangement of gunwale stones, it proved possible to locate three in the archaeological investigation. A further investigation was undertaken at the site in 2017, when a trench measuring c. 18 x 13 m was placed across the middle of the ship setting (fig. 6). This demonstrated a good agreement between the excavated stone traces and those evident from the geophysical surveys (fig. 12). During the investigation in 2017, several metal objects were found by metal detector in the soil layer located over the central part of the ship setting. These were primarily melted fragments of gold and bronze objects, as well as a silver arte­fact and one of iron. The gold objects comprise a fragment of band-like gold foil with animal ornamentation (fig. 14), two small rings of beaded wire (fig. 15a-b), a small band-like ring (fig. 16), a fragment of an edge fitting of gold foil (fig. 17), two partially melted fragments of lugs (?) (fig. 18), four unidentifiable fragments of gold foil and ten melted gold lumps. The bronze finds comprise a strap buckle (fig. 21), two possible fragments of brooches (fig. 19 and 20), a fragment of a bronze arte­fact with vaulted upper surface (fig. 22), six unidentified fragments and 44 melted bronze lumps. The silver object is a small rivet, and there is a small fragment of a thin iron sheet. As is evident from the above description, most of the metal finds from the ship setting are heavily fragmented and partially melted, which makes dating difficult. The finds are interpreted as remains from a cremation grave that has been placed on the surface of the central part of the ship setting. The most remarkable find is the ornamented gold foil (fig. 14), which can be assigned to the second half of the 6th century AD; the early part of Salin’s style II. As for the two gold rings (fig. 15), similar rings are employed on brooches and swords from the Late Germanic Iron Age, where they are placed around rivets. The third gold ring (fig. 16) must, due to its tiny diameter, be perceived as decoration that has encircled an object. The strap buckle (fig. 21) is of a type that Mogens Ørsnes dates to phases 1 and 2, i.e. c. AD 550-725. The possible button bow from a button-bow brooch (fig. 19) cannot be identified more closely. Ørsnes classifies button-bow brooches into types E0-6, which collectively can be assigned to the period AD 550-880. The fragment of an equal-armed brooch (fig. 20) cannot, due to its small size, be identified more closely within Ørsnes’ types F1-4, which are all assigned to his phase 1, c. AD 550-650. The most recent research dates the beginning of the dating range for this brooch type to between AD 510 and 545. All in all, the finds can be placed within the second half of the 6th century AD, as also suggested by the ornamented gold foil. A central question is whether the ship setting and the cremation grave are contemporaneous. With regard to the ship setting, a sample was taken for OSL dating from one of the socket stones in A121. The date subsequently obtained by the Nordic Laboratory for Luminescence Dating/DTU Nutech was 1.33±0.13 ka, i.e. 688±130 in calendar years. In other words, the ship setting was constructed in the period AD 558-818. When the OSL date is compared with the archaeological dating of the artefacts, it seems very probable that ship setting and cremation grave are contemporaneous. A further indication that the now vanished grave had a direct link with the ship setting is the distribution of the metal finds: These were all found over the central part of the monument. Burial customs in the Late Germanic Iron Age are characterised by local differences, and cremation graves surrounded by stone settings are the dominant form in Jutland. The best-known example is Lindholm Høje at Nørresundby, with ­almost 700 graves marked by various kinds of stone settings, including some that are ship-like in form. At Høj Stene there was nothing to indicate that the cremation had taken place directly on-site. All the evidence suggests that the cremation grave was located so high up that it was disturbed by the cultivation activities of later times. The presumed cremation grave shows some similarities with the situation in Grydehøj at Lejre. The base of this large round barrow, which had been c. 40 m in diameter and up to 4 m in height, was covered by a burnt layer containing charcoal, fragments of burnt bone and artefacts in the form of gold wire, drops of melted gold, melted bronze and burnt iron rivets. These represent the remains of a cremation grave containing the remnants of exceptionally rich grave furnishings that were destroyed in the cremation and, judging by patches of fire-reddened earth, this took place directly on-site. Grydehøj has been 14C dated to the first half of the 7th century AD and is therefore contemporaneous with the early magnate’s settlement at Lejre, with the Høj Stene grave, with similar cremation graves and with closely related boat graves in Sweden that, in addition to magnificent grave goods, were also accompanied by remains of numerous animals. As for the many unanswered questions about monumental ship settings, the local­isation of Høj Stene and the associated investigations have taken us a significant step further towards a better understanding of the chronology and function of these structures. The enormous ship setting is dated to the Late Germanic Iron Age, c. AD 600, and is so far the first example that can be securely assigned to this period. Furthermore, it was constructed as a sepulchral monument for an unusually richly-furnished cremation grave which exhibits many common features with the coeval elite grave in Grydehøj and with Swedish burial monuments. Like other monumental ship settings, Høj Stene occupied a striking location. It was placed by the Gudenå river, which was presumably a major traffic route extending far into the central part of Jutland. The monument must have been impressive seen from the river, and it was also sited close to an ancient traffic junction. Kongensbro lies about 1 km to the southwest, and there has been an actual bridge there since the end of the 14th century, but it could very well have been the location of a crossing place prior to this. Høj Stene’s northern and southern barrows cannot be dated more closely, but they lie together with two other barrows that probably date from the Late Neolithic Single Grave culture or the Bronze Age. It seems remarkable that large ship settings, which had not been built since the Late Bronze Age, suddenly turn up again after a break of a millennium in a totally different cultural context with a very different world of ideas. It is a well-known cultural-historical phenomenon that new and striking monuments sometimes appear formed in the shape of familiar structures from another time. In the case of Høj Stene and other contemporaneous ship settings, these should probably be interpreted as an expression of a contemporary elite’s need clearly to manifest its power. With the emergence of the Frankish Empire, which by AD 600 encompassed large parts of the Continent, Europe had gained a new superpower following the dissolution of the West ­Roman Empire. The northern border of the Frankish Empire ran close to present-day Denmark, and it became important for the people who lived in the lands to the north. Both in Scandinavia and the British Isles petty kingdoms emerged in the 6th and 7th centuries, which mimicked the Frankish warrior aristocracy. Rich burials, such as Sutton Hoo in southern England and related boat burials at Vendel and Valsgärde in Sweden, bear witness to close dynastic connections. The settlement at Lejre, with its large farm complex, the burial in Grydehøj and possible coeval ship settings, are examples which demonstrate that just such an elite existed in eastern Denmark during the 6th century. It is in this perspective that we perceive Høj Stene. We have demonstrated here the largest known ship setting from the Late Iron Age, together with the remains of heavily burnt grave goods reflecting a rich burial. There is therefore much to suggest that we are on the trail of a possible Jutish elite residence equivalent to that in Lejre. Indeed, it may already have been found. In metal-detector surveys on the fields to the north and northeast of Høj Stene in recent years several localities have been recorded with finds from the Late Germanic Iron Age (fig. 24). Seen in relation to the ship setting, the northernmost of these is of greatest interest. At this locality, which lies 2 km from the ship setting, 39 artefacts that can be assigned to the Late Germanic Iron Age have been found within an area of c. 2 ha (fig. 25-27). Moreover, there is also waste from bronze casting and slag from iron smelting. As pointed out by Birgitta Hårdh, clear evidence of high-quality metalworking is found at South Scandinavian central places such as Uppåkra in Scania and Gudme on Funen. A similar situation is suggested by the finds from the site at Borre Skov, with indications of bronze casting and iron smelting as well as a patrix (punch), which bears witness to exquisite craftmanship. The finds are contemporaneous with the ship setting, and they perhaps represent a coeval elite residence.Høj Stene A monumental ship setting by the Gudenå river Ship settings, or stone ships, in the sense of stones arranged in a regular pointed-oval shape, are known especially from Denmark and Sweden, where some can be dated to the Late Bronze Age, others to the Late Iron Age and Viking Age. Of the latter, so-called monumental ship settings stand out by virtue of having a length in excess of 40 m (fig. 1). A further general feature is that, either singly or in groups, they occupy prominent positions in the landscape, usually by a main traffic route and often in association with barrows; some are found in conjunction with a royal residence – Jelling and Lejre. Insofar as rune stones form an integral part of a ship setting, as is said to be the case at Glavendrup and Bække and is suggested by the circumstances at Jelling, the monument can be dated to the Viking Age. In other cases, the dating of the monuments is less certain, with several possibly deriving from the end of the Late Iron Age. There is also great uncertainty about the function of these monumental ship settings. There has been a general tendency to perceive them as burial structures on a par with the other ship settings. But the preservation conditions at the archaeologically investigated monuments have been poor, and it has therefore not been possible to demonstrate a clear link between monumental ship settings and possible graves. Even so, these monuments can clearly be perceived as a power manifestation by an aristocratic environment, and they can be viewed in conjunction with the ship symbolism that is evident in Late Iron Age boat graves and the spectacular ship burials of the Viking Age. With the aim of addressing these issues, Moesgaard Museum has investigated the site of one of southern Scandinavia’s largest ship settings, located at Vejerslev, close to Gudenå river (fig. 1, no. 1). The monument was demolished more than 150 years ago. The first mention of it in written sources is in 1683 under the name “Højs Steen”, located on the western fields of the village of Vejerslev, which run down towards the Gudenå. A written source from 1768 states that the ship setting extends between two small barrows standing about 100 paces apart. A report submitted to the National Museum of Denmark around 1850 contains valuable information about the ship setting as the detailed text is accompanied by a survey of the ship setting’s ground plan. It is stated that the monument was c. 88 m long and 13 m wide in the middle. The drawing shows 16 preserved stones. These stood 2-4 m apart and were c. 1.9-2.5 m high (fig. 4). But the stone at the northern stem/stern, which was c. 4.5 m in length, lay toppled on the northern barrow. The distance from the Gudenå river is given as just less than 100 m. A written source from 1877 states that the final remaining stones were removed in 1852. In 2014 and 2016, systematic geophysical surveys were undertaken of the area between the mound that could be the northern barrow and the remains of the presumed southern barrow. Collectively, the geophysical surveys sketched a picture of an extended pointed-oval formation between the two barrows (fig. 6). The structure had been c. 82 m long and 14 m wide at its broadest point. This concurs well with the dimensions given in the report to the National Museum around 1850. The ship setting was oriented NNW-SSE and lay almost parallel with the Gudenå, c. 130 m distant from it. It was located on the edge of a level sandy plain that rises above the riverbed. An archaeological investigation was undertaken in 2016, with an excavation trench measuring c. 18 x 9 m being positioned across the southern part of the ship setting, in a place where the geophysical surveys had demonstrated stone traces or possibly stones (fig. 6). Of the four anomalies evident on the geophysical surveys that fitted with an arrangement of gunwale stones, it proved possible to locate three in the archaeological investigation. A further investigation was undertaken at the site in 2017, when a trench measuring c. 18 x 13 m was placed across the middle of the ship setting (fig. 6). This demonstrated a good agreement between the excavated stone traces and those evident from the geophysical surveys (fig. 12). During the investigation in 2017, several metal objects were found by metal detector in the soil layer located over the central part of the ship setting. These were primarily melted fragments of gold and bronze objects, as well as a silver arte­fact and one of iron. The gold objects comprise a fragment of band-like gold foil with animal ornamentation (fig. 14), two small rings of beaded wire (fig. 15a-b), a small band-like ring (fig. 16), a fragment of an edge fitting of gold foil (fig. 17), two partially melted fragments of lugs (?) (fig. 18), four unidentifiable fragments of gold foil and ten melted gold lumps. The bronze finds comprise a strap buckle (fig. 21), two possible fragments of brooches (fig. 19 and 20), a fragment of a bronze arte­fact with vaulted upper surface (fig. 22), six unidentified fragments and 44 melted bronze lumps. The silver object is a small rivet, and there is a small fragment of a thin iron sheet. As is evident from the above description, most of the metal finds from the ship setting are heavily fragmented and partially melted, which makes dating difficult. The finds are interpreted as remains from a cremation grave that has been placed on the surface of the central part of the ship setting. The most remarkable find is the ornamented gold foil (fig. 14), which can be assigned to the second half of the 6th century AD; the early part of Salin’s style II. As for the two gold rings (fig. 15), similar rings are employed on brooches and swords from the Late Germanic Iron Age, where they are placed around rivets. The third gold ring (fig. 16) must, due to its tiny diameter, be perceived as decoration that has encircled an object. The strap buckle (fig. 21) is of a type that Mogens Ørsnes dates to phases 1 and 2, i.e. c. AD 550-725. The possible button bow from a button-bow brooch (fig. 19) cannot be identified more closely. Ørsnes classifies button-bow brooches into types E0-6, which collectively can be assigned to the period AD 550-880. The fragment of an equal-armed brooch (fig. 20) cannot, due to its small size, be identified more closely within Ørsnes’ types F1-4, which are all assigned to his phase 1, c. AD 550-650. The most recent research dates the beginning of the dating range for this brooch type to between AD 510 and 545. All in all, the finds can be placed within the second half of the 6th century AD, as also suggested by the ornamented gold foil. A central question is whether the ship setting and the cremation grave are contemporaneous. With regard to the ship setting, a sample was taken for OSL dating from one of the socket stones in A121. The date subsequently obtained by the Nordic Laboratory for Luminescence Dating/DTU Nutech was 1.33±0.13 ka, i.e. 688±130 in calendar years. In other words, the ship setting was constructed in the period AD 558-818. When the OSL date is compared with the archaeological dating of the artefacts, it seems very probable that ship setting and cremation grave are contemporaneous. A further indication that the now vanished grave had a direct link with the ship setting is the distribution of the metal finds: These were all found over the central part of the monument. Burial customs in the Late Germanic Iron Age are characterised by local differences, and cremation graves surrounded by stone settings are the dominant form in Jutland. The best-known example is Lindholm Høje at Nørresundby, with ­almost 700 graves marked by various kinds of stone settings, including some that are ship-like in form. At Høj Stene there was nothing to indicate that the cremation had taken place directly on-site. All the evidence suggests that the cremation grave was located so high up that it was disturbed by the cultivation activities of later times. The presumed cremation grave shows some similarities with the situation in Grydehøj at Lejre. The base of this large round barrow, which had been c. 40 m in diameter and up to 4 m in height, was covered by a burnt layer containing charcoal, fragments of burnt bone and artefacts in the form of gold wire, drops of melted gold, melted bronze and burnt iron rivets. These represent the remains of a cremation grave containing the remnants of exceptionally rich grave furnishings that were destroyed in the cremation and, judging by patches of fire-reddened earth, this took place directly on-site. Grydehøj has been 14C dated to the first half of the 7th century