Bendstrup – en fyrstegrav fra den romerske jernalder, dens sociale og historiske miljø

                     Bendstrup - A Princely Grave from the Early Roman Iron Age: Its Social and Historical Context In 1869 a 'princely grave' from the Early Roman Iron Age was discovered in a barrow on Bendstrup mark, the town of Bendstrup, Hvilsager parish, Sønderhald district, Randers county. Some of the grave goods were purchased for a private collection at the manor Valbygård near Slagelse, where they are still to be found (fig. 3-6). Through a recently discovered exchange of letters between the seller/finder, Sophus Andreas Bergsøe, and the buyer, the owner of Valbygård, August Villads Bech (nephew and uncle, respectively), it has been possible to reconstruct the find as a whole from among the many objects of the Valbygård collection (fig. 1, 2, and 7). The grave was probably an inhumation grave; the many clay vessels mentioned but not preserved indicate that this type must be classified under the common type of pottery grave of Central and Northern Jutland. It was found in a barrow beneath a pile of large stones. The orientation was said to be north-south. The grave goods preserved consist of two silver fibulae with twisted gold wires, 4 cast lion's claws, 3 small round "bases", and a "bowl" lacking a bottom, all of bronze, together with 4 fossils, 2 ammonites and brachiopods. A large amber bead could also be identified as part of the find. Together with these objects there was a quantity of pottery, and a considerable accumulation of rust on the outer side of the bronze "bowl" must come from a rather large iron object, possibly a knife, which had broken off after excavation. However, it must be noted that Sophus Bergsøe first arrived at the site after the find had been dug up, as mentioned several times in the description. Thus he refers (and interprets?) the information provided by the excavator; he personally checked through the soil which had been dug up. DESCRIPTION AND DATING OF THE FIND The silver fibulae (fig. 9-11) The fibulae are single-jointed and cast, with a length of 5.5 cm and a weight of c. 25 g each. The fibulae must be placed typologically in Almgren's gr. IV: strongly profiled fibulae and subsequently in the 1st main series: with a bow-head and an openwork catch-plate. They can most closely be compared with fig. 67 (Almgren 1923, p. 34 ff). This fibula type has an eastern main distribution, concentrated in the Austrian Alpine region, Bohemia, and Eastern Germany. It occurs less frequently towards the west, and in Denmark it is not found at all, according to Almgren. Chronologically it ought to be ascribed to the transition between the Late La Téne and the Early Roman Iron Age. However, the Bendstrup fibulae are in a class of their own, due to the animal heads. Early Roman Iron Age fibula types with a well-modelled animal head on the bow are known from the Celtic-Illyric eastern Alpine region, with a concentration south of the Roman Danube border (Jahn 1952, Werner 1954). Moreover, there is an isolated Danish group with a concentration on Djursland (fig. 20). The Danish examples are: Hoby on Lolland (fig. 12) and St. Grandløse, Holbæk county (fig. 13), which are of silver. All other animalheaded fibulae, with a single exception, can be attributed to one district, Sønderhald in Randers county: Ryomgård gr. 2 (fig. 14), Langkastrup gr. 14 (fig. 15), Langkastrup II, gr. B (fig. 16), Virring gr. 22 (fig. 17), and finally Torrild mark, Århus county (fig. 18). This last-named is made of silver; all those from Djursland are of bronze. The basis for the typological classification of the Bendstrup fibulae in the local environment is comprised of the individual stylistic elements of the animal heads: 1) the animal heads full face, 2) the animal heads in profile, 3) the shape of the eyes, and 4) the shape of the ears. These individual elements are illustrated in fig. 19. Upon this basis, it is possible to divide the animal heads into 4 typological groups: A,B,C, and D. The A-group includes Bendstrup (A 1) and Langkastrup gr. 14 (A 2). The heads are well-modelled with upper and lower jaws, carefully executed eyes consisting of several concentric circles and ovals which make them appear to "see.", and small "smile wrinkles" (particularly pronounced on A 1). The ears are shaped like small isosceles triangles. A 1 is of silver with a two-ply gold wire round the forehead. A 2 is of bronze and has a smooth silver wire round the forehead. The B-group consists of Hoby (B 1), St. Grandløse (B 2) and Virring (B 3). Full face, the main line is straight (or faintly concave and with a sharply marked central line). The lower jaw is missing on B3. The eyes are ornamental dots on B 1 -where there are 3 X 2 pairs on each fibula -or they lack entirely (B 2 and B 3). The ears are long, nearly right-angled triangles with no modelled effect. B 1 is of silver with a frontlet of two-ply gold thread; B 2 is completely of silver, and B 3 is of bronze. The C-group consists of Ryomgård gr. 2 (C 1) and Langkastrup II gr. B (C 2). The main line, full face, is long and curved with distinct cheekbones and a pointed upper jaw. The lower jaw is weak (or may lack entirely) on C 2. The eyes are two concentric ovals (C 1 and seemingly C 2), the ears are long isosceles triangles -C 1 with a plastic modelled effect, C 2 faintly engraved. Both fibulae are of bronze, C 2 quite corroded and therefore with rather vague features. The D-group consists of the silver fibula from Torrild mark. The: animal head, full face, has the same form as that of the B-group and the lower jaw is missing. The eyes are distinct, with small amorphous bumps, and the ears are extended with convex sides and a strongly modelled effect. An attempt to interpret the typological groups within the relative chronological frame­work shows the following: A 1 is completely isolated with its perfectly well-modelled animal head. This must be interpreted as the earliest form, A 2 as a later imitation. For the B-group, B 3 is younger than B 1 and B 2, as the animal head on this fibula has lost its zoomorphic character to en even greater degree. The same is the case with C 2 in relation to C 1. On the basis of the same criteria, D must be placed late stylistically. Summary: Comparison of the stylistic result, based on the analysis of individual elements, with the typological classification of the fibulae (after Almgren 1923) yields the following 3 chronological groups: La Téne III/Early Roman Iron Age: Bendstrup; 1st century AD: Langkastrup, Hoby St. Grandløse, Ryomgård gr. 2; first half of the 2nd century: Virring gr. 22, Langkastrup II gr. B, Torrild mark. The fibula types mentioned are -with the exception of type 67- common North European forms (Almgren 1923). Torrild mark type 109/110, however, is most common on Bornholm. Type 67 of the Bendstrup fibulae, however, is concentrated in the eastern Alpine region. In Northern Europe, it is only sporadically represented (Almgren 1923 p. 36 + map, Klindt-Jensen 1950 fig. 122 and p. 193). Fibulae with an animal head on the bow are found especially in the Roman provinces Noricum and Panonia and further south to the Adriatic coast around Acquileia (Werner 1954 ab. 3). In a European context, fibulae with an animal head on the bow can be divided into two distribution areas: a South European and a Danish (fig. 20). The South European type is attributed chronologically to the transition to the Early Roman period, that is, the beginning of the 1st century AD -the local Danish particularly in the second half of the 1st century and the beginning of the 2nd century (Jahn 1952 p. 97). The Bendstrup fibulae thus naturally belong to a South European context from the beginning of the 1st century AD and were, in contrast to the later Danish examples -not made locally in the North. Bronzes The Lion's Claws (fig. 21-24): Four nearly identical lion's claws are cast in bronze. Along with the modelled lion's claw, each foot is equipped with a pair of wings with palmettes. The right-angled shape shows that the lion's claws had been fastened to a square base. Remains of soft-soldering may still be seen on the upper side. On its back, each foot bears a T-shaped fabrication stamp. Three of the feet are numbered in Roman numerals from I­III; the fourth lacks a number. The weight is about 75 g for three of the feet, whereas the fourth (marked II) weighs 65 g. This foot is in other ways identical to the others, but the outermost palmette and the bottom of the paws was later ground(?) away (compare photograph, lower right) Bases (fig. 24-25). Three small cast bases of bronze have an identical surface diameter of 2.6 cm. The sides are profiled, and the diameter at the bottom is 3.2-3.5 cm. The weights vary between 26-34 g. No doubt these pieces served as bases for the lion's claws (compare photograph). A fourth base must have been lost. Bronze "bowl" (fig. 26-29). The round bronze "bowl" is also cast. Its maximum diameter is 14.3 cm, its height 4.3 cm, and the hole in the middle has an interior diameter of 6.0 cm. The exterior is convex and profiled. The interior reveals traces of casting and solder drippings. The weight is about 450 g. The bronze is badly oxidized, except for the upper edge, where the surface of fracture appears completely new. The exterior also bears a trace of rust, which must have been caused by a larger iron object lying next to the bronze bowl in the earth. Interpretation: There is no doubt that the bronzes are Roman work. In a reply to Sophus Bergsøe in the 1870s, the keeper at the Louvre, A. de Longperier, paralleled the lion's claws with similar ones from Pompeii, thus dating them to before 79 AD -a comparison and a dating which cannot be contested. All the bronzes are interpreted as connecting parts of a Greek-influenced Roman work from the beginning of the 1 st century AD, made at a workshop in central Italy. In the last century it was not possible to determine with certainty the type of bronze object to which the pieces belonged. Today, the comparative material is more copious and thus the possibility of identification is greater. Moreover, it has been possible to subject the bronzes to microscopic analysis, which makes for an even more precise interpretation. There is no doubt that the lion's claws had been soldered to a square base. The comparative material therefore consists of Roman braziers and wine craters. In both cases, the lion's claws also have bases (Pernic 1925 p. 37 f., Macchioro 1913 p. 460). An interpretation of the bronze "bowl" as the foot of a vessel of Egger's type 94 or 98 (Eggers 1951) is logical. However, the bowl can also be linked to the wine crater, as mentioned, in which case it would have been placed at the joining of the vessel to the pedestal, where it would have covered the soldered joint. If all the bronzes belonged to one and the same bronze object, the possibilities are limited to the early Roman (Pompeiian) wine craters (fig. 30). In this case it would be the first time such a piece has been documented in Germania Libera. The conservator C. Gottlieb of the National Museum in Copenhagen has subjected the bronzes to a microscopic analysis which supports the suggested interpretation. It has been determined beyond doubt that the "bowl" was not the foot of a bronze vessel, as the "support" (the maximum diameter) bore only traces of secondary wear which occurred after the period of burial in the earth. It is equally certain that the outwardly-curved edge of the "bowl" (the minimum diameter) completely lacks a patina -and thus that while it was buried in the earth it was soldered to another object (traces of soldering can be discerned). However, it has not been possible to ascertain whether the right-angled surfaces of fracture on the lion's claws, like the traces of soldering, were first made after burial in the earth. The relationship between the three small bases and the lion's claws is equally unclear, but solder is found on the upper side of the bases, and on the bottom of the lion's claws. We may thus conclude that the "bowl" was not the foot of a bronze vessel and that the theory that it was part of a wine crater accords with the completely undamaged support which must be ascribed to the placement of the "bowl" as a connecting link between crater and pedestal. We can at the same time conclude that in any case the crater itself must have been present when the Bendstrup grave was found in 1869. The fossils The four fossils consist of 2 brachiopods (fig. 8) and 2 ammonites (fig. 31-32). The ammonites can with certainty be identified as a pair from the same shell. A more exact identification of family and species cannot be made; the geological dating therefore extends from Jurassic-Cretaceous. As there seem to be no similar pieces in the Scandinavian Jurassic-Cretaceous deposits, they are more likely from German deposits, for example, from the southern Baltic, Holstein, Mecklenburg, and the Teutoburg vicinity. The function of the 4 fossils is unknown, but their presence in pairs may suggest that they were used for a game. Games of dice were common throughout Greece as well as in the Roman Empire, where they were often played with small regular bones -astragali- which have, in fact, been found in several Early Roman Iron Age graves in Jutland. The principle of the game required that the backs of these small bones be different from the fronts. The Bendstrup ammonites, which meet this requirement, are completely polished from wear. Whether or not they were used for games, the fossils add an interesting perspective to the Bendstrup find. Such worthless trifles would hardly have reached Denmark through trade or exchange through several links. They suggest a more direct relationship between the southern provenience of the find and the person interred at Bendstrup. Amber (fig. 8 and 33) One amber bead and a piece of unworked amber are mentioned by Sophus Bergsøe. Of these, only the amber bead can be identified with certainty. It has a diameter of 2.5 cm, a maximum thickness of 0.6 cm, and a perforation of 0.5 cm. A narrow groove runs along the outer edge. Unpreserved objects Many clay vessels and potsherds were reported at the excavation of the find. Some of these had retained their shiny black surface. The find also included a large iron object, to judge by the accumulation of rust on the exterior of the bronze "bowl". Dating of the find The Bendstrup find can be dated by the two silver fibulae and the bronze crater. The fibulae (Alm. IV: 67) are assigned to the transition from La Téne III/Early Roman Iron Age, that is, the close of the 1st century BC/beginning of the 1st century AD; and the bronze crater is dated to the beginning of the 1st century AD. Through the grave type and the grave goods, the find may be definitely assigned to the beginning of the Early Roman Iron Age. Of course, there is always some degree of uncertainty with regard to when the objects were made in relation to when they were buried. But this uncertainty is not particularly great, as the fibulae bear no traces of wear. THE SOCIAL CONTEXT Introduction The theoretical basis of the following analysis is the view that socially dominant groups who represent major productive conditions will seek to legitimize their dominance through an ideology bearing upon society as a whole, and expressed through symbols, social norms, rules, and rituals. This ideology is not considered to be a passive reflection of the society but, on the contrary, an active factor used by competing individuals and social groups to establish and legitimize their dominance by making their principle generally accepted in the societal organization. This implies that the ideology may entail both contrast and harmony on various levels -for example, among competing social groups or within various activities in the society, such as the use of the agricultural production, either for basic reproduction or for the establishment of political alliances through ceremonies, the exchange of gifts, etc. For example, if the production drops below the level necessary for basic reproduction, it would threaten the politics of alliance of the dominating groups, which are based upon a surplus, and consequently social tensions would arise which in the final end could trigger a gradual restructuring of the society. The social, economic, and ideological structures thus modify one another in a dynamic interplay, whose catalysts are embedded in the social and political/ideological structure, but whose potential and limitation lie in the economic structure and the ecology. This dynamic principle can ideologically be manifest either in a reinforcement or a repression of various aspects of the society, depending upon its complexity and stability. For example, wealthy persons of high rank may be interred modestly, with but few symbols of their wealth and position, whereas rising and competing social groups may be buried with a great symbolic manifestation of their social position. Frequently, only a single social group or segment of the society expresses its status in burials, whereas in some periods quite the opposite is the case: a considerably larger part of the population is buried, and the internal differences are sought equalized instead of emphasized. At present we know only very little regarding the societal factors which cause these variations in mortuary practices, and indeed whether general rules governing these variations can be established. In studying mortuary practices, it is therefore essential to clarify the relationship between the social reality and its ideological interpretation. Many factors are entailed here, and in the following we shall analyse some of them on the basis of the theory that mortuary practices first and foremost serve a purpose in the living society. Combinations of grave goods and cultural identification Most Danish parallels to the Bendstrup fibulae are to be found in the great cemeteries in Randers county. Four cemeteries have yielded the same number of graves with animal­headed fibulae (fig. 14-17). Together with identical features in the grave type, these graves show a striking similarity in the composition of their grave goods: pottery, an iron knife, silver/bronze fibulae, bronze pins, S-shaped hooks, glass beads, large amber beads, and sometimes unworked amber. The Bendstrup grave contains objects which correspond to the combinations in other graves containing animal-headed fibulae. The group of graves with animal-headed fibulae includes three others: Hoby (fig. 12), St. Grandløse (fig. 13), and Torrild Mark (fig. 18), but these graves do not contain the combination type mentioned. The combination of pottery and one or more iron knives is the most frequently encountered among the Early Roman Iron Age graves in Denmark. However, there are very few with the combination of fibulae (silver or bronze), pins (silver or bronze), S-shaped hooks (commonly silver, sometimes bronze), small glass beads (most often light-green and/or milky-white with gold foil), and amber bead(s), (sometimes large beads/rings). In Denmark as a whole, this combination occurs in 7 cases (group A). The combination of only 4 elements occurs in 12 cases (gr. B); and with 3 elements, in 24 cases (gr. C). (compare fig. 34). Their distribution shows a combination of grave goods which is characteristic of a small group of graves at the large cemeteries without tumuli in Århus and Randers counties, and whose main distribution moreover is in Eastern Jutland (distribution, fig. 35). The relationship of this group to the other graves in the area will be investigated next. The comparative material covers all graves from Randers, Århus, Skanderborg, and Vejle counties, together with adjoining districts in Ringkøbing and Viborg counties (compare map, fig. 36). Within this area, attention is focused on graves with "rich" grave goods including: weapons and/or spurs, imports, berloques, and gold rings. The result of these investigations can be summarized in a geographical model (fig. 44) in which each circle represents the geographical relationships observed. Graves with the grave-goods combination A-C and graves with the weapon combination I-III constitute the two main groups in the region (weapon combination I: sword, shield, spear, and sometimes spurs and/or an axe; II: sword and shield or spear and shield; III; sword, spear, or shield). Each of the two groups has a "core area": the one consists of the rich combination in the ornamental equipment (typ. A-B) the other of the well-equipped weapon graves (typ. I + spurs). Where the two groups adjoin geographically there is one (double?) grave with contents from both groups. The Roman bronze imports are linked especially to the area with the weapon graves of typ. I. The most significant observation is that core areas with the richest combinations are mutually exclusive. The area with animal-headed fibulae and grave goods of groups A-B are complementary to the area with the richly furnished weapon graves (typ. I + spurs), as well as to the region with gold rings. Two areas, a southern and a northern, are particularly distinct and differ greatly from one another. The southern group is concentrated especially in the region between Horsens and Kolding. The concentration of graves is far less than to the north (see map, fig. 37) and this is the only county in the investigation area in which cremation graves predominate (c. 75% - see fig. 45). Richly furnished weapon graves and spur graves are found throughout the region, and graves with imported goods are nearby. However, women's graves with the grave­goods combinations A-B lack completely, and graves of the C-type occur only sporadically. The cemeteries are smal