Abstract:Mesolithic settlements by the lakes and watercourses of the Gudenå systemIn the late 19th and early 20th century, several amateur archaeologists collected considerable quantities of archaeological artefacts from a large number of Stone Age settlements in Jutland, especially along the upper reaches of the river Gudenå at Tørring. In the 1930s, these activities were made more systematic, and supplemented by field-walking, minor excavations and collection along the major river and lake systems by archaeologist Therkel Mathiassen. The collective results of these investigations were published as Gudenaa-Kulturen (The Gudenå culture) in 1937. On the basis of comparative studies involving the Mesolithic cultures of eastern Denmark, S.H. Andersen and N. Sterum (1971) argued that the localities containing artefacts from the Gudenå culture actually represented mixed assemblages and were therefore of little scientific value. Their conclusion has, in many ways, had a deterrent effect on the last five decades of research into Mesolithic inland settlements in Jutland.Since 2011, six Mesolithic localities have been excavated within Museum Silkeborg’s archaeological jurisdiction:Holm Mølle was encountered by excavation on the planned route of a high-voltage power cable (figs. 2-4). Excavation revealed a c. 30 m2 area containing worked flint together with a small charcoal-rich hollow surrounded by oval greyish features. In and around this presumed hearth lay a quantity of white-burnt flint and charred hazelnut shells.Sølund comprised a bleached sand layer containing finds that covered an area of c. 1500 m2 and yielded exclusively flint from the Maglemose culture (figs. 5-6). At the base of this sand layer, thin charcoal layers were observed, which contained burnt flint and charred hazelnut shells. The site is of considerable extent compared to other excavated coeval localities in central Jutland.Sølyst. Excavation revealed about 500 m2 of flint-containing cultural layers that – with the aid of microliths – were dated to the Maglemose culture (figs. 7-8). The layers were more complex than first realised, and the presence of several springs had led to the formation of extensive peat cover, which sealed finds-rich layers up to 50 cm in thickness.Sørkelvej proved to be an extensive stratified cultural layer, with distinct layers containing tools from, respectively, the Maglemose and Kongemose cultures, cut through by later features, including a Neolithic pit (figs. 10-11). This is the first time that a stratigraphic sequence has been demonstrated at an inland site in central Jutland, where material from the Maglemose culture is directly overlain by material from the Kongemose culture.Sørkelvej II. Adjustments to the project at Sørkelvej resulted in further investigations in the area, leading to the discovery, beneath extensive peat layers, of a minor flint concentration around a small charcoal-rich patch (figs. 12-13). The flint assemblage is dated to the Maglemose culture.Kirkebjerggård III. At the edge of a small wetland area, occasional flint concentrations were encountered in association with an earlier vegetation surface in the bog (fig. 14). The finds, some of which are made from frost-shattered flint, include several scrapers, a core borer, a hammerstone and two pressure flakers/fabricators.With time, the results of these new investigations can hopefully contribute significantly to a better understanding of the Mesolithic record in Jutland, as well as counteracting some of the reluctance to deal with the sites here, after the so-called Gudenå culture was discredited and dismissed in 1971.The 90 localities recorded in total within Museum Silkeborg’s archaeological jurisdiction are very largely represented by surface collections. In general, they are situated in the immediate vicinity of watercourses and lakes. The major localities all lie by open water, in the form of lake and river systems, and presumably reflect repeated occupations over time. A few sites are though situated on higher ground, between 70 and 530 m from the nearest watercourse or wetland area.The recently excavated sites open up new perspectives for further investigation of Mesolithic inland settlements and the possibility of encountering in situ structures and even stratigraphy. It seems therefore likely that there is excellent potential for productive investigation of several other “old” sites that have been more or less dismissed as mixed surface assemblages.The recently excavated Mesolithic sites confirm the picture formed earlier with respect to a classic location in the immediate vicinity of the lakes and rivers of the Gudenå system. The preliminary findings suggest that the major localities lie by lakes or largewatercourses, while the smaller localities, representing a single brief occupation, are generally found further inland. It is remarkable that, despite shifting climatic conditions, people in the Stone Age repeatedly returned to the same localities by lakes and rivers. We can possibly perceive these sites as “enduring places” that, in addition to their ecological-economic value, also had cultural significance.The major localities appear to be the products of numerous individual events, some of which overlap horizontally and have thereby resulted in relatively significant quantities of finds distributed over extensive areas at these sites.The artefact assemblages dating from the Maglemose culture can, with time, contribute to identification of the Jutlandic flint inventory and yield important information on the exploitation of the Jutland interior – also with respect to resources located further away from the major watercourses and lakes.Despite the fact that the stratigraphy at Sørkelvej appears to support a cultural development – reflected in the microliths – that follows the established eastern Danish sequence, this development has not been elucidated satisfactorily in the Jutland record. In northern Germany, sites from the Kongemose culture are virtually absent, and much now suggests that core elements of the Kongemose culture are not of local origin, but arrived here from the east. Furthermore, around 6400 BC, a marked change took place in the hydrology of Denmark, which can be linked to the so-called “8.2 kiloyear event”. The latter affected the weather, climate and environment in the northern hemisphere. These climatic changes took place more or less simultaneously with the massive Storegga tsunami, which directly affected Mesolithic settlements in for example Norway and the British Isles. Its effect has also been demonstrated recently in Denmark, despite the fact that the majority of the coastline affected at that time now lies submerged beneath the sea. Significant changes can be traced in the settlement patterns and the material culture of this period in several areas of northern Europe, when climatic and environmental events, together with possible migrations and changes in the contact network of the time, led to innovation in and transformation of the tool inventories. It is striking in this respect that none of the newly-discovered sites has an absolute date later than 6400-6200 BC, the time when coastal localities become common in the Danish archaeological record. Even though there are many uncertain factors involved, this could possibly reflect a shift in settlement strategy, from a fundamentally inland orientation to a subsequent coastal orientation. The data foundation is still too slight for absolute conclusions to be drawn with respect to the internal versus external developmental influences, and how these relate to for example the chronological developments at the Stone Age localities. The long-term aim is to be able to place recently excavated sites in a broader national and international (northern European) context and use them to formulate new archaeological and environmental working hypotheses for future investigations in the region.Karen Rysgaard Kaj Fredsgaard Rasmussen Silkeborg MuseumFelix RiedeAarhus Universitet Institut for Kultur og Samfund Afdeling for Arkæologi

Abstract:Bronze Age cult houses in Thy Thy is renowned for its many burial mounds from the Bronze Age. During the last 150 years, the mounds have yielded numerous finds from both the early and the late Bronze Age. Naturally, the finds from inside the mounds, from the graves, have been in focus. However, recent investigations have made it increasingly obvious that the graves are sometimes just part of the story. In several cases, structures closely related to burials inside the mounds have in fact been found outside them, often built right up to the foot of the mound. In the context of the Danish Bronze Age this is a rather new realization, as demonstrated by the following presentation of cult structures from the late Bronze Age in Thy. It is a common feature of these structures that traces of them can be found next to large ploughed-over burial mounds, where the levelled filling of the mound has completely or partly protected them against destruction caused by ploughing (Fig. 1).The first well-documented structure of this type emerged more than twenty years ago outside a large, ploughed-over burial mound near Thisted in northern Thy, called “Høghs Høj” ( “høj” being the Danish word for mound). The mound had a diameter of 25 metres and was originally constructed over an inhumation grave from period III of the early Bronze Age (Fig. 2, structure N2). However, the mound also contained secondary burials, such as an urn grave from period IV of the late Bronze Age (Fig. 2, structure N5) dug into the foot of the mound towards the southwest. A number of stone-paved areas were uncovered just outside the urn grave, built right up against the row of kerb stones (Figs. 2-4). These stone-paved areas consisted of a single or double line of stones surrounding a square area of approximately 6 by 6 metres (Fig. 2, structure N19). Towards the north, another stone paving, separated from the first one by a stone-free area 1.5-metre wide (Fig. 2, structure N17), had a number of oval stone-free areas marked out (Fig. 3). These stone-free areas may indicate the position of posts, although no post-holes in the subsoil confirmed this theory. No other post-holes connected with the structure were found. Part of the construction furthest away from the foot of the mound has been destroyed by modern ploughing. However, the narrow outer stone paving probably continued in a curve, as strongly indicated by marks in the subsoil (Fig. 3).Connected to the stone-paved areas, and inside the square structure, lithic debitage and some late Bronze Age pottery, probably from period IV or V, were found. The finds do not differ from those known from contemporary Bronze Age settlements, yet the connection with the stone-paved areas and the mound placed the finds in a context somehow connected to the grave cult. This interpretation has been confirmed later by finds of a similar character outside burial mounds in Thy and Mors (Fig. 1). Most of those finds, however, had been severely destroyed by ploughing. Apart from Grydehøj and Gramstrup near Vestervig, the three best parallels are from ploughed-over mounds at Nørhå and near Sundby in central and eastern Thy and at Toftum in southern Thy (Figs. 1, no. 5, 6, and 12). In all three cases, the remains were found of a stone-paved area adjoining the foot of the mound, and in each case, one or more secondary graves from the late Bronze Age were found in the mound just behind the stone-paved area. Figs. 5 and 6 show the finds from Toftum and Sundby. Just as was the case at Høghs Høj, the location of the secondary graves is hardly accidental. At Toftum, two out of three graves positively date from period IV, and the Sundby mound contained an urn grave from period IV.Until now, the best example of a cult construction outside a mound is from Grydehøj in southern Thy (Fig. 1, no. 10), where it can definitely be said that an actual cult house was erected in the late Bronze Age. The mound turned out to originate from the single grave period, but it had been extended during the early Bronze Age. Three secondary cremation graves were found just inside the kerb stones, dug into the Bronze Age mound (Figs. 7, 5-7). One was a small stone cist with just a few burnt bones in the filling; the other was a cremation grave, which must have held a small wooden coffin that contained a little heap of cleaned, white-burnt, crushed bones. The third grave was an urn grave with some white-burnt, crushed bones. None of the graves contained datable grave goods. In Thy, the small stone cist and its surrounding gravel filling would normally date this grave to the first half of the late Bronze Age. The two other graves cannot yet be dated.During the first half of the late Bronze Age, two or more structures were erected outside the line of kerb stones on the southern side. Apparently, one replaced the other (Fig. 7, 8-9). The oldest and best-preserved structure is described in the following as a cult house (fig. 7, 9), which was later partly covered by a completely different, ramp-like structure (Fig. 7, 8). Other stone-paved areas lay towards the east, but they were too poorly preserved to allow any essential conclusions to be drawn about their appearance and purpose (Fig. 7, 10).Once the ramp-like structure was fully investigated, the older structure was revealed (Fig. 8). Its central part was a house with an inner dimension of 5.5 by 5.0 m. It consisted of a wall ditch with traces of rather stout posts placed at relatively regular intervals. Towards the south, the house had an approximately 1-m wide and 1.5 to 1.7-m long, funnel-shaped entrance, and outside the easternmost entrance ditch, a dug hole must have held a post or a stone. West and east of the house were narrow stone-paved areas, seven and four metres long respectively, and around 1.25-m wide. Unfortunately, ploughing had destroyed part of the stone-paved areas, which may have continued in front of the house to the entrance. Before the house was built, the terrain underneath and around it seems to have been levelled. This meant that material was added west of the house and removed east of the house, so that the house was constructed on an approximately horizontal plot. This was revealed by the fact that the western wall ditch of the house had been dug into clayey turfs that had been placed on end, and this layer continued under the long, narrow paving towards the west. Under the eastern paving, in contrast, the subsoil had been removed. The just over 1.5-metre wide area between the wall ditch and the narrow areas of paving had held a thick turf wall. This wall had been preserved to a height of 30 to 40 cm between the back wall and the row of kerb stones defining the foot of the mound. Towards the south, the turf wall seems to have continued to the entrance, which explains why the entrance was so long. Ditches were also found between the surrounding narrow stone-paved areas and the turf wall. They probably contained posts which were intended to prevent the heavy peat wall from sliding onto the paving. The stone-paved areas were either exposed or functioned as the base of an earth layer. Perhaps they should be interpreted as procession paths. No artefacts were found underneath, inside, or on top of the turf wall, whereas pottery and flint were found both on the surrounding stone-paved areas and inside the house. Numerous stones found inside the house must be the remains of a structure (Fig. 8). This could not have been a floor, as more often than not the stones were situated on top of the other finds in the house. The distribution of the stones inside the house was concentrated in the areas along the walls, especially the well-preserved corners beside the back wall, with a concentration descending towards the centre of the house. This indicates that they could be the remains of a levelled stone bench originally built along parts of the wall.A single, quite deep, post-hole was found in the centre of the house, underneath the stone layer and the floor layer. A post placed in this hole must have carried the roof of the house (Fig. 9, 4). Further in, closer to the back wall of the house, the bottom of a fireplace appeared in the floor layer as an area of reddish-brown burnt clay (Fig. 9, 5). The fireplace was situated at the end of one of two curved ditches which formed a semi-circular construction or enclosure built up against the back wall of the house. The opening in this enclosure was positioned directly opposite the entrance of the house. This construction probably represented the sanctum of the house. The finds from within the house consisted almost entirely of thin sherds from good-quality pottery. Most of them came from relatively small vessels, probably goblets. A preliminary dating of the sherds dates the house to period IV or V of the later Bronze Age. Most of the pottery was found just inside the entrance and in the back third of the house, on both sides of the entrance to the semi-circular enclosure (Fig. 10). A small amount of white-burnt broken bones found in the house has been investigated, but whether they were from humans or animals was not determined. If they are human, these bones may constitute a direct connection to one or more of the cremation graves situated just behind the house. It is therefore possible that for reasons still concealed to us the burnt bones were placed in the house before they were buried.Quite a few cooking-pits, apparently contemporary with the cult structures, were found outside the house and the narrow stone-paved areas (Fig. 7). The cooking-pits outside and the drinking vessels inside the house indicate that ceremonies involving eating and drinking formed part of the grave cult. The ploughed-over mound of Gramstrup is situated only 900 m northwest of Grydehøj. The investigation of this mound began in the same year as the Grydehøj excavations finished. Quite unexpectedly, a cult structure was also found near the foot of this mound. Besides having a number of similarities to Grydehøj, the Gramstrup structures provided opportunities for new observations. As in the Grydehøj case, the mound was erected during the single grave period. All in all, this mound had six or perhaps seven phases, the last of which – involving enlargement of the mound to a diameter of c.25 m – was probably not constructed until period II or III of the early Bronze Age. There were no preserved grave-finds from that time, however.The Gramstrup cult structure measured 8 by 13 m. As was the case at Grydehøj, the structure was flanked by narrow stone-paved areas adjoining the eastern foot of the mound (Figs. 11 and 12). The paved areas consisted of stones of a diameter of 10 to 15 cm, arranged in a single layer. The best preserved northern paving, which had a width of 0.5 to 1.0 m and a length of 6 m, continued at a right angle for another couple of metres into the area in front of the structure towards the east, as was probably also the case at Grydehøj. The southern stone paving probably had a similar course originally, but it was poorly preserved. The outer edge of the northern stone paving was made up by a row of somewhat larger stones, carefully arranged (Fig. 13). The distance between the outer edges of the two stone-paved areas was approximately 13 metres. This and other measurements were very similar to what was observed at Grydehøj. However, the Gramstrup structure had traces of a somewhat different construction inside the stone-paved areas. In the Grydehøj case, we were dealing with a square ditch structure, interpreted as an actual house, whereas in Gramstrup there were traces from a circular structure some 6.5 to 7-m in diameter, presenting itself as a border of flat stones preserved in situ, and as stone impressions (Fig. 12,2). The stone border had a width of up to 1 m, and because of its circular course, it was tempting at first to interpret it as a chain of kerb stones surrounding a completely vanished small mound, which had been built up against the Bronze Age mound. This theory, however, is contradicted by a number of circumstances – for instance the complete lack of mound-filling within the circle, and the fact that the stone border is placed exactly in the middle of the symmetrical axis of the entire cult structure. Moreover, the diameter of the circular stone border almost completely matches the inner diameter of the Grydehøj cult house. For these reasons, there is no doubt that the structure should be considered an integrated part of the cult structure; it is an entirely different matter, however, to establish whether or not a house was erected on this site, as was the case at Grydehøj.On the inside of the stone-paved areas (Fig. 12, 5), exactly as at Grydehøj, a c.1.5-m wide turf wall had been built. The turf wall was observed as being stratigraphically later, both to the north and to the south, than some collapsed filling from the mound. This is in keeping with the fact that we are dealing with a structure built onto the already existing mound. Inside the turf wall, the circular stone border may be regarded as some sort of bench or seating construction. The distance from the outer stone border to the turf wall is at least 0.5 m. The shape towards the east, where the entrance to the whole structure must have been, is unknown, as not many traces are left from the stone border in this area. As the turf structure resembles that of Grydehøj, it is possible that a cult house also existed next to the Gramstrup mound. This interpretation is supported by the fact that a post in the centre of the structure (Fig. 12,14) may have supported the roof, and that a few post-holes, perhaps from an inner partition wall, were observed on the inside of the northern turf wall, between the wall and the stone border. In the area between the turf wall and the surrounding stone paving – within the wall, that is – traces were found from a number of rammed-down posts, which probably supported the wall on the outside (Fig. 12, 3). The distance between these posts indicates that they were joined together by interlacing branches. As opposed to the Grydehøj case, no artefacts were found in the central part of the structure. Only a small concentration of pottery sherds of uncertain date was found on the northern stone paving. A strong blade-knife of flint with a retouched back, of a type usually dating from the late Bronze Age, was found close to the northern stone paving.The only Bronze Age grave found in the Gramstrup mound was dug into the mound-filling about 1 m west of the back of the cult structure (Fig. 12, 1). The structure was probably made in connection with this burial. The grave was a cremation grave in a trough-like pit measuring 1.2 by c.0.75 m. At either end there was a rather large stone resting on the bottom of the grave. Between the stones, and about half way down the filling, was a thick layer of burnt, crushed bones. The layer covered a small area of the grave pit measuring approximately 33 by 60 cm and had a depth of 5 cm. As is the case with many late Bronze Age graves in Thy, the grave filling consisted of cleaned pebbles, in this case beach stones worn by being rolled by water. The results of a C-14 dating of the bones are not yet available. The strong likeness between the Gramstrup grave and two of the Grydehøj graves indicates that they are probably closely contemporary. A dating of these graves to period IV of the late Bronze Age fits well with the dating to the transition between period III/IV or period IV of a somewhat larger, but structurally very similar male grave at Vibberstoft, Villerslev parish, which had been placed at the edge of an older mound (cf. Fig. 1, no. 7). Here, the remains of badly damaged stone-paved areas were found outside the grave and up against the foot of the mound. They are interpreted as the remains of an almost completely destroyed cult structure. If a line were drawn through the centre of the cult structure at Gramstrup, through the middle of the two flanking stone-paved areas, and through the centre of the stone border within these, it would hit the middle of the grave in the foot of the mound precisely. This cannot be a coincidence, and it confirms the connection between the grave and the cult structure, as known from other similar structures in Thy.The cult houses from Grydehøj and Gramstrup have provided us with a key to explaining a number of other structures in northwest Jutland – first of all, the structure found at Høghs Høj, which may now be interpreted with certainty as a cult house almost identical to the one found at Grydehøj. Even when it comes to size, the square inner structure from Høghs Høj (Fig. 4) matches with a surprising degree of similarity the inside of the Grydehøj house (Fig. 9). As for the Sundby mound and other finds of stone-paved areas outside mounds in Thy and on Mors (cf. Fig. 1), for the time being we must merely note that they are related to the Grydehøj house, although there is no guarantee that a house was indeed built next to these mounds. Stone-paved areas alone may have indicated a ceremonial area in front of the mounds and may not have been connected with any building.In Kobberup near Skive (Fig. 1, no. 16), an almost circular structure was excavated. It was built up against a mound with a megalith chamber as its primary grave (Fig. 14). In this particular case an inner, semi-circular enclosure had been built onto the older mound, like the examples in Thy. Behind it, inside the foot of the older mound, were two possible cremation graves. This find has both striking differences from the finds in Thy and features that strongly resemble them, but unfortunately the dating of the structure is uncertain.From the rest of Denmark, parallels to the cult structures mentioned here are still few. One find from Zealand is rather similar, though. In Ballermosen near Jægerspris in Hornsherred a small rectangular building built up against the foot of an early Bronze Age mound was interpreted by the excavator as a cult house. In spite of the differences between this find and those from NW Jutland, there are strong indications that the house in Ballermosen was connected to the grave cult in the same way that the houses of the Grydehøj type were. In the case of another site in North Zealand, the house at Sandagergård, there cannot be said to be a connection similar to the one established in Thy between the mound and the cult structure, but the cult house and the contemporary graves within the house there were most probably related. It is this connection between graves – with or without a mound – and cult structures that is no doubt the decisive element, and it can be seen in different versions also in Sweden and Northern Germany. Although the new cult structures from Thy seem to be of a local design, they should no doubt be seen as part of a larger context.There is much to indicate that structures outside the mounds were in fact extremely common, and that at a certain period they were even the rule rather than the exception. The Thy finds have really made evident in earnest the prospects opened up by this fact, as is illustrated by a new structure uncovered next to Høghs Høj.In 2000, the remains of a new cult structure were uncovered next to a hitherto unknown demolished mound, immediately southeast of Høghs Høj (Fig. 15). Although this find is poorly preserved compared with the ones excavated earlier, there is no doubt that this is a structure of the same category. One could hardly wish for a better illustration of the common nature of this type of structure.Bjarne Henning NielsenVesthimmerlands MuseumJens-Henrik BechMuseet for Thy og Vester HanherredTranslated by Annette Lerche Trolle

Multi-Channel Ground-Penetrating Radar Array Surveys of the Iron Age and Medieval Ringfort Bårby on the Island of Öland, Sweden

Improved High-Resolution Gpr Imaging and Characterization of Prehistoric Archaeological Features by Means of Attribute Analysis

Integrating Airborne Laser Scanning and 3D Ground-Penetrating Radar for the Investigation of Protohistoric Structures in Croatian Istria

A multi-proxy geochemical and micromorphological study of the use of space and stratigraphy of a Viking-age house in Ribe, Denmark

Temporal dynamics of Linearbandkeramik houses and settlements, and their implications for detecting the environmental impact of early farming

GPR Imaging and Characterization of Ancient Roman Ruins in the Aquileia Archaeological Park, NE Italy

Unveiling buried aeolian landscapes: reconstructing a late Holocene dune environment using 3D ground‐penetrating radar

Monitoring Erosion of a Submerged Stone Age Site and Its Implications for Managing Underwater Cultural Heritage. A Case Study at Tybrind Vig, Denmark

Stone-Paved Cellars In The Stone Age? Archaeological Evidence For A Neolithic Subterranean Construction From Nygårdsvej 3, Falster, Denmark

Mesolitiske bosættelser ved Gudenåsystemets søer og åer

Mobility patterns in inland southwestern Sweden during the Neolithic and Early Bronze Age

Mellemholm

Mobility among the stone age island foragers of Jettböle, Åland, investigated through high-resolution strontium isotope ratio analysis

Landscape development at Lina myr fen, Eastern Gotland, 9000−2500 cal. yr BP

JORDFALLET AT BOHUS: REINTERPRETING THE 14C DATING OF A MEDIEVAL LANDSLIDE EVENT

Sr analyses from only known Scandinavian cremation cemetery in Britain illuminate early Viking journey with horse and dog across the North Sea

UAV LiDAR in coastal environments: Archaeological case studies from Tierra del Fuego, Argentina, and Vega, Norway

Thickness Estimation of the Soil‐Sedimentary Cover Inside Causewayed Enclosures to Locate an Occupation Layer: Map of the Archaeological Potential of the Neolithic Causewayed Enclosure of Le Pontet

Multi‐Method Structural Investigation of the Schneiderberg–Baalberge Burial Mound (Saxony‐Anhalt, Germany) Including Seismic Full‐Waveform Inversion (FWI)

Intensively Cultivated Roman Villae Estates: Case Study of Medulin Bay (Istria, Croatia)

Bronzealderens kulthuse i Thy – Anlæg med relation til gravkulten