Decorated Standing Stones – The Hagbards Galge Monument in Southwest Sweden

5.1 The Stone Settings

Circular stone settings are a well-known archaeological phenomenon in Scandinavia and have a wide dating from the Late Bronze Age to the Late Iron Age (1100 BC–1000 AD). They appear both individually without any accompanying graves, in smaller groups together with other burials, and at gravefields (Carlie, 1994, pp. 61–71).

Until now the discussion on Hagbards galge has focused on the standing stones and rock art, while the stone settings themselves have been left out of the discussion. The well-preserved stone setting is 24 m in diameter and stands out from other examples due to its size. According to a compilation made by Bo Strömberg, which is based on data from Fornsök, there are only two stone settings that are larger in mid-Halland (Strömberg, 2005, Figures 169–172). Looking into the more local area, north of the city of Halmstad and south of the River Ätran, Hagbards galge is the largest stone setting (Strömberg, 2005, Figures 170 and 171).

Monumentality is often regarded as a Bronze Age feature (Lundborg, 1972, 2007), which seems to be confirmed by excavated stone settings in mid-Halland. Those dated by archeological methods to the Early Iron Age fall between 3 and 14 m in diameter, and stone settings dating to the Late Bronze Age fall within a similar range, but they also include three large burials 19, 20, and 23 m in diameter (Strömberg, 2005, Figure 173). Based on these considerations it seems reasonable that the larger and monumental stone setting at Hagbards galge was constructed in the Late Bronze Age 1100–500 BC, even though alternative datings cannot be excluded.

5.2 The Standings Stones

In Scandinavia, standing stones are a well-known archaeological phenomenon and have a general dating to the Iron Age (500 BC–1000 AD). They are most often restricted to grave fields, although they can also occur individually. In general, these stones are 1–2 m tall (Carlie, 1994, pp. 81–85).

Based on this, the stones at Hagbards galge are comparatively very tall, three are 4 m tall, and one is 3.5 m tall. This height is not unique as a couple of taller standing stones exist in other parts of Halland (Table 2). However, in mid-Halland, there is only one other stone of a similar height, and it is situated only 1 km to the southeast (Fornsök: L1998:8267).

As noted, standing stones are generally dated to the Iron Age (500 BC–1000 AD) and it is not rare to see this statement as an unproblematized fact (Carlie, 1994, p. 85; Hyenstrand, 1984). However, it has long been known that there exists a group of standing stones that can be assumed to have a Bronze Age dating due to their association with dated burials (Ericsson, 2005; Forssander, 1940). Often the dating is uncertain since it is not the standing stones themselves that are dated, but a nearby monument. In the case of Hagbards galge, the rock art indicates that the standing stones had a use phase during the Bronze Age. That the stones were already standing upright during this phase is indicated by the fact that at least one of the stones had rock art on all sides including the one that was originally attached to the bedrock (see Broström & Ihrestam, 2016a). In addition, the directionality of all the carvings complies with the standing stones, something that would not necessarily be the case were the stone still more or less flat and attached to the bedrock. Finally, there is no vivid rock art tradition on panels in Halland meaning that we cannot assume that there were any rock art panels available to be quarried and used as stelae.

To get a more precise idea of the dating of standing stones within this period, information was gathered on standing stones in excavated Bronze Age contexts in the province of Halland. The results are accounted for in Table 3.

The sample displayed in Table 3 is very small and consists of only 11 standing stones at 10 sites. In one case, the stone is freestanding and dated by pottery found in a nearby pit to 800–400 BC. In two cases, stones were put on top of excavated mounds. One of these mounds yielded pottery that could be dated to the transition between the Bronze Age and the Iron Age (Harplinge; L1997:971), and the other contained pottery of a more general Bronze Age dating (Vinberg; L1996:551). While there are some uncertainties concerning the relationship between the standing stones and the pottery found nearby, the situation is different in those cases where standing stones were found below mounds.

For example, in 1971, a mound that was 26 m in diameter was excavated in Snöstorp outside Halmstad. A 2.3 m tall stone was found lying down 3 m inside the edge of the mound. Originally, it had been situated in a stone foundation which was documented close to the stone. In the mound, there were two burials dating to Montelius’ period 5 (900–700 BC), and there was an additional plundered grave in the center of the mound (Lundborg, 2007, pp. 19–25). A likely scenario is that the mound was originally smaller and accompanied by a standing stone c. 2 m tall. In period 5 (900–700 BC), the stone was overturned, and the mound enlarged. The dating of the standing stone should thereby be period 5 or earlier.

Standing stones have also been documented in a couple of other mounds, though their height was only one meter or less.

In conclusion, the data regarding standing stones in regional Bronze Age contexts are limited, but the information indicates that standing stones were also raised in the Bronze Age and not only in the Iron Age. Interestingly, the oldest securely dated standing stones in the province appear in Bronze Age period 4, 1100–900 BC. However, it should be noted that by widening the geographical perspective the pattern gets more complicated as standing stones in other parts of Sweden have also been dated to the Early Bronze Age. These include older examples where the information is somewhat insufficient (Forssander, 1940), but also new examples, such as a 1.1 m-high standing stone at Sommaränge skog in Uppland situated in a layer including burned human bones dating to the earliest Bronze Age (Forsman & Victor, 2007, pp. 60–64).

5.3 The Rock Art

A newer documentation using the tactile tracing method was carried out by Sven-Gunnar Broström och Kenneth Ihrestam in 2015 (Broström & Ihrestam, 2016a). During this survey, they identified three human figures on one face of the stone 17:3 (Figure 5 top) (Fornsök: L1998:8552). On another face of the same stone, concentric circles seem to have been accompanied by two lines below the circles which were curved at the lower ends. One of the ends extends to the top where it terminates with an equally curved ending indicating perhaps another human figure. Most of the results of the examination in 2016 were confirmed by our documentation. However, the lower limits of stone 17:3 were not reached because the rock was too wet to make the targets for the laser scanner stick (Figure 5).

Figure 5

Visualization from a laser scan documentation of the rock art using TVT (https://tvt.dh.gu.se/); (a) stone 17:3 and (b) stone 17:6. Laser scan: Rich Potter; visualization: Christian Horn.

Some differences in recordings were however identified. It was previously uncertain whether the figure on stone 17:6 (Figure 5, bottom) (Fornsök: L1998:8254) could be interpreted as a warrior. The discovered “hooks” on the two parallel lines below the concentric circles can be interpreted as feet, while a single line above the circle may have represented a head with an extended neck and nose or beak. The carving can therefore be interpreted as an anthropomorphic figure with a shield. All these characteristics have parallels, for example, the large warrior with spear and shield in Finntorp (Fornsök: L1968:7100), Bohuslän, albeit there is no spear on the figure in Hagbards galge.

From the second outer ring, a wide and comparatively deep line extends. This line was identified in the 1980s by Bergström and Lundborg who documented it as a line without any further interpretation (Bergström & Lundborg, 1985). The line is not mentioned by Broström and Ihrestam and is rather smooth compared to the roughness of the gneiss. Bergström and Lundborg are probably right in identifying it as a carved feature. On its upper end, it appears that several semi-circular depressions were placed next to each other with slight overlaps. That the lower end of the line is mostly within the line of the circle also speaks for an intentional placement. Judging by its form, the line could be the carving of an individual sword.

If this was a carving of an individual sword, then it provides details that can be interpreted. The hilt has several pronounced swellings and an inconspicuous hilt terminus which could be compared to the so-called Dreiwulstschwert type that derives its archaeological name from such features (Figure 6). Another name for these swords (Scheibenknaufschwerter) derives from their simple discoid hilt terminus. Similar swellings can be observed, for example, on Late Bronze Age antenna swords. However, these stand out because of their hilt terminus (see Laux, 2009, Pl 33). Two Dreiwulst-swords have been discovered in Denmark, i.e., Borbjerg and Ringkøbing, which both had very worn-down hilts (Thrane, 1969). There is a lengthy debate as to the dating of these swords, which we cannot discuss fully here so we will only use the most common datings. Thrane (1969) saw the imported swords in Denmark based on typological characteristics as belonging to the Hallstatt (Ha) phases A1 and A2. Later, Sperber maintained that these swords began during the Central European Bronze Age phase D, but he agrees that most of them date to the Hallstatt (Ha) phase A1 (Sperber, 1987). If the sword at Hagbards galge depicts a Dreiwulst-sword, then the engraving was made at the earliest c. 1200–1050 BC. In relative terms, this is the late period 3 transitioning to period 4, or the transition from the Early to the Late Bronze Age.

Figure 6

Comparison between the hilt of the Dreiwulstschwert from Borbjerg, Denmark (a), the reconstructed hilt (b), and the hilt of the rock art sword (c). All scaled to the same size (drawing by Christian Horn, a. after Thrane, 1969).

The shield of the human figure itself only has ribs without decorative buckles. The closest parallel is the shields of the Harlech type (Figure 7a, b, and e) which has its main distribution in south-east Britain, with no physical finds in Scandinavia. While it could also be another simple type like the Nipperwiese type, this form has less ribs than depicted on stele 17:6 and is, therefore, a less well-suited comparison (Figure 8). The Harlech type also begins in Bronze Age phase D and declines at the beginning of Ha A1, around 1200 BC (Uckelmann, 2012).

Figure 7

Comparison between (a) the shield from Harlech, (b) a reconstruction and simplification of the shield from Harlech, (c) the shield from Lommelev, (d) a reconstruction and simplification of the shield from Lommelev, and (e) the shield from stele Asige 17:6. All scaled to the same size (drawn by Christian Horn, (a) after Uckelmann 2012, Tab. 20, (c) after Uckelmann, 2012, Tab. 2).

Figure 8

Comparison between (a) the shield from Nipperwiese, (b) a reconstruction and simplification, and (c) the shield from stele Asige 17:3. All scaled to the same size (drawn by Christian Horn, (a) after Uckelmann, 2012, Tab. 14).

Other shields could also have been depicted if we assume that some artistic liberties were taken in depicting them, or that there were more variations in real life than are archaeologically known. One such aspect could be that the many buckles some shield types display were ignored by the carvers here. As such the slightly larger distance between the outer two circles could indicate a shield of the Lommelev-Nyírtura type (Figure 7c–e). This would establish a direct connection between Central Europe and the Scandinavian sphere because physical finds have been discovered in both regions. However, the difference in the distance between the two shield ribs may be coincidental, and in addition, it would not change the dating of the engraving (Uckelmann, 2012).

Although rock art typically depicts metalwork weaponry, the shield engraving could of course be a representation of organic shields such as those discovered in Ireland (Uckelmann, 2012, pp. 72–73). These finds demonstrated that presumably well-dated shield types, in this case, shields of the Herzsprung type (which have never been found in Halland) may have had very similar-looking predecessors that date much earlier (Uckelmann, 2012, pp. 72–73). Should any shield type – organic or not – have been depicted on the stele that we do not know archaeologically, then there is no chance of dating the engravings.

Assuming we have identified the typological parallels correctly, it must be noted that the types of the sword and the shield depicted at Hagbards galge remained contemporary for a long time, meaning that the shield could have been carved only a short time before the sword, though this is at the limits of the chronological resolution for the Nordic Bronze Age.

On stone 17:3, the laser scan also confirmed all discoveries made by Broström and Ihrestam (2016a; see also Bergström & Lundborg, 1985), but it was possible to add more detail (Figure 5, top). On a human figure seemingly striding toward the right on the older documentation, a nose or beak and a foot on the left leg were discovered indicating that the figure walks toward the left. It was suspected that the concentric circles could have been another anthropomorphic figure. This could not be verified but a potential additional line was found underneath the circles. The figure on the top may have a phallus and a sword, although this area is affected by some exfoliation. Furthermore, a nose or beak was noticed, which means the figure was looking toward the right. On the anthropomorph below this figure, the lower right leg including a foot and the right upper arm were identified. The left arm is affected by a cupmark or damage.

A major new discovery is that a line identified in 2015 is actually the prow of a boat. The keel line was affected by some irregularities in the rock, but the rest of the petroglyph is clear enough. There also seems to be an extension to the right, and while it may also be extended to the left, the roughness of the rock and some exfoliation make it difficult to be certain. The outward turn of the prow dates the boat to the Late Bronze Age (Ling, 2014). The chronological relation to the other motif could not be established because of a lack of clear superimpositions. On another face of this stone, two more ships were discovered in 2015. However, these remain incomplete and very simplistic. At least the upper of these two boats could date to somewhat earlier times because it seemingly lacks an out-turned prow (Ling, 2014).

It could be argued that the concentric circles at the center of the stele represent a shield of the same type as on 17:6. However, the lower number of ribs could indicate a shield of the Nipperwiese type, which would be a better parallel (Figure 8). This type is most densely distributed in Central Europe, with a few occurrences in south-eastern England. The dating for this type perhaps already begins at the end of period II, about 1300 BC, but it is more securely placed and has its main phase simultaneously to the Harlech-type shields (Uckelmann, 2012) meaning that the shields on both stelae were perhaps carved at the same time. The boat on 17:3 is very fragmentary and difficult to date, but the prow may depict a horse head which would suit the dating of the other carvings well. Overall, it can be said that while there may be more than one carving episode on the stelae, they seem to be limited to a relatively short period between 1200 and 1000 BC.

It has been claimed that there are two ship images on the standing stone Asige 17:7 (Fornsök: L1998:7944) (Bergström & Lundborg, 1985), but this was not confirmed by Broström and Ihrestam’s documentation.

The rock art documented at Hagbards galge stands out as rather unique from a regional perspective because figurative rock art is rare in Halland. Even though new documentation of rock art has been carried out in recent years, there is no obvious parallel to Hagbards galge (Broström & Ihrestam, 2016b, c, 2017a, b, 2018). The dominating images are footsoles appearing on cliffs and boulders, but in two instances, rock art has been identified on standing stones. These stones are rather small, measuring 1.5 and 1.0 m in height and the rock art consists only of parallel carved lines, a small circular image, an unidentified figure, and some cupmarks (Broström & Ihrestam, 2016b).

Also, on a national level, it is difficult to find close parallels to the carvings at the Hagbards galge monument. The Klintastenen at the island Öland in the Baltic was found in a mound measuring 18 m in diameter in 1915. The stone is 1.2 m high and decorated with a crewed ship, a circular image, and four horses (Arne, 1917). Another possible parallel is the Villfara stone found in 1863 in a mound in southeast Scania. The stone is 1 m high and engraved with 2 ships, a horse-drawn wheeled vehicle, a horse, 23 cup marks, and 3 unidentifiable motifs. The dating of both the Klinta stone and the Villfara stone is Montelius’ period 2 (1500–1300 BC) (Johannsen, 2013). The dating of these monuments together with the size and context of the decorated stones make these a less convincing comparison to the Hagbards galge monument.

6.1 Chronology

Given the monumental size of the stone setting dating to the Late Bronze Age is likely (1100–500 BC). This assumption is further underlined by the standing stones, which appear in different Bronze Age contexts supported by regional data which indicates that these kinds of stones appear in the Late Bronze Age (1100–500 BCE) rather than in the Early Bronze Age (1700–1100 BCE (Table 2). This conclusion is also supported by a comparative northwest-European perspective. In Britain and Ireland standing stones appear in a variety of archaeological periods including the Bronze Age (O’Sullivan & Downey, 2020; Williams, 1988). However, excavations reveal that Bronze Age standing stones, including both individual and paired standing stones, have a rather limited chronology appearing from c. 1200–900 BC (Higginbottom, 2020; O’Brien, 2023). Interestingly, the newly discovered Dreiwulstschwert-type sword image fits well into such a chronology, since if the attribution of the sword type is correct, it would date to 1200–950 BC (Sperber, 1987). The dating of the either Harlech-type or the Lommelev-Nyírtura-type shield also fits into this framework (Uckelmann, 2012).

If the stone setting, the standing stones, and the rock art were made during a limited period of time, dating to late Montelius’ 3 or Montelius’ period 4 (1200–900 BC) seems most likely. However, we cannot be sure that all the components making up the monument, i.e., the stone setting, the standing stones, and the rock art were made at the same time. For example, the stone setting could have been constructed first, and at a later stage, the standing stones were added. Whether or not the rock art was originally part of the standing stones is an open question. It is also not known how many carving episodes are represented by the rock art. Given these uncertainties, it is perhaps more relevant to think of Hagbards galge as a site with an extended biography which probably gained its present shape in the period 1200–900 BC.

6.2 The Communicative Landscape

The rock art images, revealing a combination of a type of sword and shield, which are hitherto unknown in Scandinavia, speak in favor of the buried person having belonged to a European elite milieu. Something which is further underlined by the monumental standing stones and the impressive size of the best-preserved stone setting. To understand how such an elite milieu developed in Mid-Halland, we must consider the monument’s position in the local and regional landscape context, as well as within a network on a European level, facilitating the movement of people and goods.

Starting with the regional setting, it is obvious that the monument held a strategic position in relation to contemporary transportation routes. As noted, Hagbards galge is situated along an ancient road dotted by mounds, testifying that it was already in use during the Bronze Age (Strömberg, 2005). Thus, the group that controlled the Hagbards galge monument monitored the movement of people and goods going through Halland in a north–south direction. The centrality of this place is further underlined by the nearby enclosure with a possible Bronze Age dating.

Furthermore, there is a relationship between the Hagbards galge monument and traffic to and from the province of Västergötland. This area could be reached from three different directions. One was following the River Göta Älv upstream into Västergötland (Nimura et al., 2020) and the other was following the inland routes along the River Viskan or the River Ätran (Strömberg, 2005, p. 310; Weiler, 1994). For people using an inland route to travel from the metal-rich areas in present-day Denmark and Scania in southernmost Sweden, the Ätran route and would have been the best option. Traveling this way, one would pass by Hagbards galge before continuing to Västergötland. Based on the distribution of metal artifacts, the province of Västergötland was a wealthy and major center both in the Early and Late Bronze Age (Larsson, 1986).

Even though the Hagbards galge monument is situated inland, it is only 10 km from the outlet of the River Suseån, which is a natural harbor in this part of Halland. Nearby there is a deposition of a golden bowl and a sword both dating to period 4 (1100–900 BC) that marks the importance of this place as a hub in a network facilitating the movement of both people and goods (Nimura & Skoglund, 2024).

6.3 The European Backdrop

The presumed dating of Hagbards galge to late Montelius’ period 3 or Montelius’ period 4 (1200–900 BC) places the monument in a time when Scandinavian Bronze Age societies were re-organized after a period of crisis, which could help to explain the emergence of the site. Judging from the abundance of metal artifacts, Montelius’ period 2 (1500–1300 BC) represents the peak of the Early Bronze Age (1700–1100 BC).

This situation changed in Montelius’ period 3 (1300–1100 BC) when there were interruptions to the supply of metal from the Continent toward Scandinavia, affecting the hoarding frequency which fell compared to the previous period (Larsson, 1986, p. 176; Levy, 1982, p. 60) and is also evident in the province of Halland (Skoglund et al., 2024). Another sign of crisis in Montelius’ period 3 is that the circulation time for swords was prolonged in certain regions (Kristiansen, 1978; Kristiansen & Suchowska-Ducke, 2015). During this period, there was a gradual change in the concentration of wealth from the area around Schleswig-Holstein toward south-east Scandinavia (Kristiansen, 1978, p. 168), and regions like south Halland (Lundborg, 1972, 2007) and south-east Scania (Håkansson, 1985) witnessed an increase in the number of burials and associated metal artifacts. These changes were related to a major re-organization of the trade networks on the continent (Horn et al., 2024; Kristiansen & Suchowska-Ducke, 2015). Emerging out of this situation were new trade networks that linked the Baltic toward the Carpathian basin through the mouth of the River Oder (Kristiansen & Suchowska-Ducke, 2015).

Around 1100 BC a new and more stable situation emerged, and metal now arrived in Scandinavia from two major routes: one from the south leading to the Carpathian basin (Melheim et al, 2018) and the other from the west including the British Isles and Iberia (Ling et al., 2014; Ling et al., 2019). Interestingly, both these trade networks – Britain, the Carpathian basin, and Iberia – are reflected in the rock art: the depicted sword of the so-called Dreiwulstschwert-type originated in Central Europe while the depicted shield is of the Harlech type and has its main distribution in south-east Britain. Furthermore, the combination of sword and shield is a frequent occurrence on Iberian stelae, for example, El Carneril/Trujillo or Torrejón Rubio 1 which provides an intriguing parallel to the standing stones at Hagbards galge (Díaz-Guardamino et al., 2019). Thus, the imagery on the standing stones at Hagbards galge situates those responsible for carving it within a code of a warrior ideal that was recognized in many European regions and relied on depicting warriors and their paraphernalia (Díaz-Guardamino et al., 2022).

As stated in the introduction there are several burials in south Scandinavia and northern Germany that stand out due to their impressive size and richness of finds, and are often labeled as “princely graves” or “kings graves.” These graves include the Kivik burial in southernmost Sweden dating to 1500–1300 BC (Goldhahn, 2009, 2013), Hågahögen in eastern Middle Sweden dating to 1100–900 BC (Almgren, 1905; Eriksson, 2022a; Ullén & Drenzel, 2018), Luseøj in Denmark dating to 900–700 BC (Thrane, 1984), and Seddin in northern Germany dating to the same period (Hansen & Schopper, 2019). Among these, the closest parallel to Hagbards galge is the Hågahögen mound in eastern middle Sweden.

The monumental mound, which is 45 m in diameter and 7 m high, was excavated in 1902–03 and several objects were recovered (Almgren, 1905; Eriksson, 2022a; Ullén & Drenzel, 2018). Among these finds were a flanged hilted sword decorated with gold, a golden fibula, a razor decorated with gold, and several other objects. The artifacts date the mound to Montelius’ period 4 (1100–900 BC). The mound is positioned not far away from a hillfort with a Bronze Age dating, a monumental cult house, and is also close to a maritime route (Olausson & Bornfalk Back, 2022; Victor, 2002). It has been hypothesized that the Håga family controlled the movement of people and goods in this part of Scandinavia to at least some extent (Eriksson, 2022b). Even though there are obvious differences between Hagbards galge and Hågahögen related to the design of the monuments, both probably date to the same period, and both are positioned strategically in a communicative landscape close to nearby enclosures or hillforts. In this, we may see the formation of a new elite phenomenon related to the control of a re-intensified trading network emerging around 1100 BC.