Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 18 Sanctuaries as Anchoring Points for Community Construction The construction of collective identities through sanctuaries and assembly places is a key element for the understanding of past societies and in particular their dynamics of aggregation. Major sanctuaries of civic religion, as well as meeting places of inter-group cult communities, offer one of the best possibilities for gaining access to political and ethnic constructs at different scales of social organization. As several historical and ethnographic studies clearly show, the regional and supra-regional cult centers must have played a key role in the formation of groups with a shared identity (Earle 1997:153). These sites would be “the concrete anchoring points in the landscape where the polity’s core values—as exemplified in its tradition of origin—were transmitted to the wider community through recitals, dramatic performances and collective rituals” (Derks and Roymans 2009:8). In other words, these would be the places where the “creation of tradition” could take place, a fundamental aspect of ethnogenetic processes (Roymans 2004, Wenskus 1961). In a world imbued by religion and that did not distinguish between sacred and profane, ritual spaces would have served as landmarks in the territory, acting as elements that brought together different extended family groups. They would therefore be sites in which politics, religion, and the building of collective identities would go hand in hand, fulfilling a fundamental role in establishing, maintaining, and strengthening ethnic ties (Gerritsen and Roymans 2006). In some way, we could say that it was around these centers that communities were “constructed”. It is important to note that supra-local sanctuaries and/or assembly places need not necessarily have been situated within central settlements; they may also have existed away from habitation or burial sites. An example of this would be the political meeting place (and central cult site?) of the Batavian elite, which according to Tacitus (Hist. 4, 14) was located in a sacred forest (Fig. 1). In this paper, we will consider two different models of central focal points from the 2nd and 1st centuries BC: on the one hand, the principal sanctuaries and public places of the Middle Rhine-Moselle region, which were situated within the major Late La Tène fortified settlements known as oppida; and on the other hand, the sanctuaries of the Lower Rhine, which were not directly linked to fortified habitation sites. Although in both cases the identification of pre-Roman sanctuaries is heavily influenced by the later building of Gallo-Roman temples on the same locations, recent research points towards an Iron Age origin of their sacred significance, something which is particularly clear in cases such as Titelberg, Martberg, or Empel (see Fernández-Götz 2014 and Roymans 2004 for a more complete discussion). It needs to be stressed that the Middle and Lower Rhine are representative of two great landscapes with different characteristics, in which different types of societies developed. From the undulating countryside of Eifel and the Ardennes to the south in the Late La Tène period, we find communities with traits such as oppida, the presence of Mediterranean products, and diversified coinage from relatively early dates. In contrast, on the plains to the north we find more decentralized societies, without oppida, with a more heterarchical landscape, few or no southern imports, the late appearance of coinage, and an overwhelming predominance of handmade pottery (Fernández-Götz 2014:143–228). Be that as it may, in both cases the crucial function of the central places was to provide an arena for politics at the level of the sub-ethnic and ethnic communities of pre-Roman Gaul (Fig. 2). These “politicized ethnic The Politics of Identity: Late Iron Age Sanctuaries in the Rhineland Manuel Fernández-Götz1,* and Nico Roymans2 Abstract - The Late Iron Age in the Rhineland area was a period of intensive social change, manifested in the development of a hierarchical system of sanctuaries. This paper discusses the social implications of this development, thereby emphasizing the role of regional and supraregional cult places as key-sites in the construction of politicized ethnic identities and associated power networks. Moreover, some interesting spatial and temporal patterns can be observed. In the Middle Rhine- Moselle area, the main sanctuaries and assembly places seem to be located in major fortified settlements (oppida) and often seem to have been the oldest elements within these sites. In the Lower Rhine region, there is no link between cult centers and fortified settlements, and at least one of the regional cult sites was situated in a forest. Debating the Thing in the North: The Assembly Project II Journal of the North Atlantic 1School of History, Classics, and Archaeology, University of Edinburgh, William Robertson Wing, Old Medical School, Teviot Place, Edinburgh EH8 9AG UK. 2Faculty of Arts (Department of Archaeology), VU University Amsterdam, De Boelelaan 1105, 1081 HV Amsterdam, Netherlands. *Corresponding author - M.Fernandez-Gotz@ed.ac.uk. 2015 Special Volume 8:18–32 Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 19 identities” are referred to by Caesar as pagi and civitates, and can be best defined as ethnic groupings functioning as political communities (Fernández- Götz 2014:Chapter 3, Fichtl 2012a). Public Spaces within Major Settlements: The Middle Rhine-Moselle Area In the Late La Tène period, the Middle Rhine- Moselle region emerged as the homeland of the Treveri, one of the main Late Iron Age Gallic polities. They were mentioned for the first time by Caesar, who referred to them in each of the eight books of The Gallic Wars. Recent work on the oppida of this area has offered extensive information about public spaces and sanctuaries within these central places, which are dated between the end of the 2nd and the 1st century BC (Fernández-Götz 2012, 2014; Krausse 2006; Metzler et al. 2006). The most extensively researched settlement in the territory of the Treveri is the Titelberg oppidum in Luxembourg, which covers an area of 43 ha (Metzler 1995). The more than 5000 “Celtic” coins documented, the evidence of Mediterranean imports such as Dressel 1 amphorae, Campanian pottery, and Italic bronze vessels attests to its considerable prosperity. There is also strong evidence of different activities involving artisan crafts on a scale that went beyond supplying local needs, notably iron and bronze working and the minting of coins. But the most outstanding feature of Titelberg is the so-called public space where assemblies, fairs, and religious ceremonies were held (Metzler 2006, Metzler et al. 2006). This was a large area sacra covering 10 ha to the east of the oppidum, surrounded by a ditch and a mudbrick wall built on a stone base (Fig. 3). The ditch, 500 m long, 4 m wide, and 2.5 m deep, was excavated in the rock. Its infill contained abundant animal bones, around a hundred brooches, and numerous other finds such as spearheads, miniature weapons, coins, and fragments of human skulls. These finds demonstrate that cult practices took place here and indicate that the boundary marked by the ditch was not only physical but also to a large extent symbolic, as it separated the sacred space from the profane. Excavations have Figure 1. Drawing of the political meeting of the Batavian aristocracy at a cult place in a holy forest. Etching by Antonio Tempestata and Otto van Veen (1612). Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 20 determined that the ditch was made around 100 BC at the same time the murus gallicus was built, which indicates that rigorous large-scale planning of the oppidum took place around that time. Given that Titelberg’s period of greatest occupation and splendor was in LT D2 (ibid.), that is, several decades later, we can conclude that a significant permanent settlement and growing economic activity seem to have developed because of its importance as a meeting place for politico-religious celebrations, not vice versa. First the sacred space for public events was defined, and then the settlement, artisanal production, and trade took off. Most of the public space remained open and free from any structures during the La Tène period, which meant it could hold a great number of people who gathered for large public meetings, like the Treveran assembly described by Caesar (see below). However, excavations in the southern section have allowed a succession of structures to be identified in the area situated at the highest point of the oppidum that has come to be called the “monumental center”. The constructions began in LT D1 and culminated in the Gallo-Roman period with a monumental fanum, which survived until its destruction in Late Antiquity. During most of the first half of the 1st century BC, rows of parallel movable palisades were built marking out corridors about 4 m wide and at least 60 m long that were arranged perpendicularly to the main road through the oppidum. These structures, which according to stratigraphic studies were erected and taken down many times, have been interpreted as voting installations similar to the saepta of Italian cities like Paestum, Fregellae, or Rome. Although this interpretation remains hypothetical, the identification of structures of this kind in a Gallic oppidum is reinforced by the discovery at Gournay-sur- Aronde of structural arrangements very similar to those at the Titelberg (Fig. 4; Brunaux et al. 1985, Fichtl 2010). Even before the mid-1st century BC, a huge 15 m x 14 m three-aisled building was erected on the traces of these voting corridors on the highest point of the oppidum. The site chosen was by no means coincidental, since the building was situated exactly on the axis created by the two gates giving access to the oppidum, and was thus immediately visible to any person who entered it. Moreover, a square was built in the approximately 40 m space between this monumental building and the main thoroughfare. In the center there was a stone altar, surrounded by large pits and various hearths of the same period. Several fragments of human skulls found in this area suggest the existence of ancestor worship. During the second decade BC, these monumental constructions were taken down and the large cult ditch filled in. This event coincided with and was probably due to the founding of the new capital of the civitas, Augusta Treverorum, which replaced Figure 2. Map with the main sites mentioned in the text. Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 21 Titelberg as the political and religious center of the Treveri (Metzler 2008). The site on which the main building was erected was covered by a paved platform dotted with numerous hearths and some light structures. A new open building, this time in stone, was built in the time of Tiberius, again on the highest point. Finally, a large Gallo-Roman fanum was erected at the same place in the 2nd century AD, surviving until it was destroyed during the invasions of the 3rd century AD. The succession over the generations of structures on the highest part of the oppidum, their location within the large public space, and the fact that the monumental program culminated in the construction of a huge Gallo-Roman temple indicates the sacred nature of the site. At the same time, the detailed study of the over 100,000 documented animal bones related to the public space provides evidence that animal slaughter was carried out on an almost industrial scale (Metzler 2006, Metzler et al. 2006). This, together with traces of occasional work in leather and bone, suggests that fairs and markets were held throughout most of the 1st century BC, probably linked to religious festivals. In short, the 10 ha of Titelberg’s large public space must have been used for purposes in which religious, political, and economic activities were inextricably bound together. It is tempting to think about general assemblies that could have been similar to the óenacha of Ancient Ireland or the Thing of the Scandinavian world. People whose daily lives were lived dispersed through the rural hinterland, had the opportunity to meet each other, exchange goods and information, establish closer social ties, arrange marriages, attend religious ceremonies, etc. It was a way of reaffirming the social order, power relations, and the sense of belonging to a wider community (Fernández-Götz 2014:Chapter 6). The extensive research carried out in the Titelberg, the exceptional character of the finds, and their Figure 3. Plan of the Titelberg oppidum: (1) cultic ditch that marks the boundary of the public space, (2) excavation of the monumental center, (3) concentration of the La Tène habitat, (4) possible Roman military establishment, (5) east gate, and (6) west gate. Re-drawn after Metzler et al. (2006). Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 22 importance in relation to identity justifies its extensive treatment here. However, it is also necessary to take a brief look at the other Treveran oppida and in particular the evidence they offer in relation to the politico-religious arena (Fernández-Götz 2012, Metzler 1991, Metzler and Scheid 2011). After Titelberg, the most important center seems to have been the oppidum of the Martberg, situated in the eastern part of the civitas at a height of some 200 m above the lower course of the Moselle. The site consists of two hilltops, Martberg and Hüttenberg, joined by a narrow corridor and which together covered a total of 70 ha. Its fame goes back to the 19th-century discovery of a votive inscription dedicated to Lenus Mars, the principal deity of the Treveri. Archaeological excavations during the last twenty years have led to a complete investigation of the sanctuary and part of the settlement (Nickel 2013, Nickel et al. 2008). Together with houses and granaries, some small areas free of constructions that might be interpreted as streets or squares have been documented, and textile production, metallurgy, coin minting, and long-distance trade including the presence of Roman amphorae, are also well attested. The settlement was occupied as an oppidum until the beginning of the Gallo-Roman period, and was then abandoned in favor of a new settlement in the valley, with the monumental sanctuary and a series of installations associated with the latter remaining on the hill. The sanctuary is situated on the highest part of the oppidum and is highly complex, as it comprises 12 phases that date from the beginning of the 1st century BC to the end of the 4th century AD (Nickel et al. 2008; Thoma 2000, 2007). The profusion of finds is exceptional, with more than 7000 coins and hundreds of brooches that have been documented in the area of the sanctuary alone. It should also be noted that numerous weapons, brooches, and La Tène coins were subject to ritual mutilation. Different items such as Nauheim brooches and coins make it possible to date the beginnings of cult activity to LT D1. From the first half of the 1st century BC onwards there was a public space, approximately 50 m x 60 m, at the highest point of the oppidum, enclosed by a palisade. It was in this area that the sanctuary developed through successive stages up to Late Antiquity. Originally, the interior remained free of buildings, but in the center there was an enclosure, 10 m x 12 m, surrounded by a ditch. This would be the heart of the sanctuary during the centuries that followed, the site of Temple K. Later, sometime between the 50s BC and the final decades of the 1st century BC, an area measuring 100 m x 103 m was established with a V-shaped ditch round it. Although the interpretation of this enclosure is not clear, both the line it follows—which cuts through the earlier public space without taking it into consideration—and some finds of militaria mean it could have been a Roman military establishment erected to monitor the politico-religious center of the oppidum. It was of a temporary nature, since the ditch had been filled in and the enclosure of the previous public space had been renewed before the end of the Augustan period. The first stone temples were built at the end of the 1st century AD and at the beginning of the 2nd, forming part of a complex that would be remodeled several times. Although it differed from the Titelberg, we should keep in mind Figure 4. Structures interpreted as voting installations from the public space at the oppidum of Titelberg. After Fichtl (2010), modified. Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 23 Early and Late La Tène periods, this area remained free of any construction throughout the Iron Age, which proves its special symbolism. Various structures have been found nearby, including a roofed pit dated to LT D1 that contained among other materials the charred bones of pigs and birds, a large quantity of indigenous pottery, some coins and brooches, fragments of Republican amphorae, and a bronze strainer, finds that point to cult activities already taking place here before the Roman conquest. No later than the final decades of the 1st century BC, a small wooden temple was built at the highest point of the oppidum. This building, along with the discovery of the votive offerings, suggests that the sacred nature of the public space dated back to pre-Roman times. The subsequent history of this place underscores this interpretation, as a Gallo-Roman sanctuary consisting of two temples was also developed there during the first centuries AD. Very little was known until recently about Kastel-Staadt, an oppidum of 30 ha overlooking the River Saar. Its location, along with the fortification that at Martberg too there had been a sacred public space on the highest part of the oppidum since at least the beginning of the 1st century BC . In general terms, this model is also repeated in Wallendorf Castellberg, which covers 41 ha (Krausse 2006:146–230). After the traces of occupation dating to the 5th and 4th centuries BC, there was a hiatus during the Middle La Tène period. The site was reoccupied at the end of the 2nd century BC. The finds indicate that between 130/120 and 60/50 BC there was a settlement at Wallendorf that acted as a regional center where metallurgical work was carried out and coins were also minted. In contrast, the scarcity of materials and structures attributable to the second half of the 1st century BC indicate a decline and at least partial abandonment of the settlement during this period. Of particular significance for macro-regional comparisons is the fact that there was a public space of approximately 60 m x 30 m at the highest point of the oppidum (Fig. 5; Krausse and Nübold 2007). Despite the development of a settlement in both the Figure 5. Plan of the areas excavated in the Gallo-Roman sanctuary of Wallendorf and its surroundings. Re-drawn after Krausse and Nübold (2007). Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 24 the top of the fortified area; he also drew attention to the fact that the Gallo-Roman temple was located exactly in the center of the resulting enclosure. Be that as it may, and irrespective of the nature and purpose of the ditch, the highest point of the oppidum also appears to have housed a place of cult significance dating back to the La Tène period. The largest fortified center in the area under study is the impressive Donnersberg, which in the Late La Tène period covered 240 ha (Zeeb-Lanz 2008). Although various finds confirm that this mountain, with a height of 687 m, was already frequented during the Neolithic and at the time of the Urnfield culture and the Early Iron Age, the fortification of the immense enclosure dates to the Late La Tène period. The oppidum was founded around 130 BC and survived until it was abandoned around the middle of the 1st century BC or even a little earlier. In total it had 8.5 km of Pfostenschlitzmauer walls, and was divided into two large enclosures separated by an internal wall. The nature of the terrain largely prevent any structures or buildings being identified, but the finds collected include, among others, pottery sherds, hand mills, iron tools, fragments of glass bracelets, Dressel 1B amphorae, and coins. The minting of coins is also attested. In addition to the size of its defensive fortifications, the most outstanding feature of Donnersberg is the presence of a 98 m x 66 m rectangular enclosure, which has traditionally been included in the category of Viereckschanzen (Fig. 6; Engels 1976, Zeeb-Lanz 2012:224–225). However, this interpretations has always raised a number of questions, because it would be the only example of this type found inside an oppidum. Although we are far from being able to establish a precise date for this large enclosure, it is quite possible, on the basis of pottery found within it, that it was constructed before the development of the oppidum. Considering its location at the top of the mountain, identifying it as a Viereckschanze creates unnecessary confusion. Both its structure and location make it more appropriate to classify this enclosure in the same group as the assembly places of Martberg and “La Terrasse” at Bibracte. A new name has recently been added to the list of oppida in the Treveran area: Bleidenberg bei Oberfell, which has an area of 18 ha and is situated on a hilltop overlooking the Moselle (Brücken 2008). Surface finds testify to the site’s occupation in the Neolithic, in the Urnfield period and, above all, during the Late Hallstatt/Early La Tène on one hand and the Late La Tène period on the other. The system of fortification displays various phases, which appear to date to both the Early La Tène and the Late La Tène periods. To date, no public space/ sanctuary has been found. But this could be due and finds of numerous “Celtic” coins have long suggested the existence of an oppidum, a hypothesis that has been confirmed thanks to research carried out in recent years (Nortmann and Peiter 2004). Numerous fragments of amphorae and iron slag testify to commercial and craft activities. But the most remarkable discovery has without doubt been that of a Gallo-Roman sanctuary situated at the highest point of the oppidum (Nortmann 2009). Various finds such as gold coins or parts of a sword show that the origins of this cult place may go back to the pre-Roman period, thus repeating the schema documented in the oppida discussed earlier. The absence of Late La Tène settlement structures in this sector could even suggest the existence of a public space of the kind documented at sites such as Wallendorf or Martberg. In the Gallo-Roman period, the sanctuary was accompanied by a cult theater with a capacity for 3000 people, many more than the resident population at that time. This feature confirms that even in the early centuries AD Kastel-Staadt maintained its pre-Roman role as a place of aggregation for the communities in its hinterland. As different scholars have pointed out, theaters in northern Gaul served as places for the populations of the pagi to gather and hold assemblies, i.e., they would primarily be buildings designed for meetings and ritual comitia (Derks 1998, Trunk 2007). The next oppidum to be discussed is Otzenhausen, a site which is remarkable above all because of its colossal wall (Hornung 2012). The site consists of a double-fortified enclosure, which covers a total of 18.5 ha. However, the chronological factor has to be remembered, since new research has demonstrated both the construction of a wall in the Early La Tène period and the existence of several phases of Late La Tène defenses. The effective surface area of the settlement was little more than 10 ha. In its interior, excavations have produced finds such as iron slag and amphorae. The settlement appears to have been abandoned around the mid-1st century BC, probably due to the Roman conquest of the territory or its repercussions. Although no evidence of a Gallo-Roman settlement has been found inside the fortified area at Otzenhausen, it should be noted that a small temple dating from the 2nd to 3rd centuries AD was built on the highest point. The great number of spearheads found nearby and the general abundance of La Tène finds discovered in the area, with some exceptional deposits such as a bronze pendant or a gold ring, suggest a pre-Roman origin for the cult tradition (Fritsch 2009, Hornung 2012). Furthermore, two stretches of ditch have been interpreted by Metzler (1991) as the remains of a dug boundary that would have delimited a public space of several hectares at Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 25 to the fact that this site was discovered only a few years ago and little archaeological work has yet been done on it. Whatever the explanation, it is significant that on the highest point of the oppidum there is a pilgrimage church of Romanic origin, a situation that reminds us of the one observed in Bibracte (Fleischer and Rieckhoff 2002). Although many questions remain, we find in the case of the Treveri a perfect example of the interrelationship between political, religious, and economic power. The application of the Thiessen Polygon Method shows a nearly regular spatial distribution of the oppida, with the Treveran civitas subdivided into six or seven entities. The territory of the polity was made up of different entities, each one with an oppidum with a sanctuary at its core. Moreover, it is interesting to note that all Treveran oppida show an earlier phase of occupation or at least occasional use and sometimes also fortification of their sites during the Late Hallstatt and/or Early La Tène periods, followed by a more or less pronounced hiatus during Middle La Tène and reoccupation in the Late La Tène period. The fact that the highest points within sites like Wallendorf remained free of residential structures in both the Early and Late La Tène periods can only be attributed to conscious choice, suggesting that their sacred significance may date back to the 5th or 4th centuries BC (Fernández-Götz 2014). The role of the Late Iron Age Treveran oppida as political and religious centers is not only confirmed by their regular distribution in space and by the presence of public spaces and sanctuaries inside their walled perimeters, but also by the discovery that at least four of them also acted as minting centers (Kaczynski 2009). Thus we can conclude that these huge fortified sites were places for assemblies (= political role), collective rituals (= religious role), and fairs and the minting of coins (= economic role). In this context, it should be noted that the best known example of a Gallic public assembly is the Treveran gathering convened by Indutiomarus: “he proclaimed an armed convention. This in the practice of the Gauls marks the beginning of a war; and by a general law all grown men are accustomed to assemble at it in arms, while the one who comes last to the assembly is put to death with every kind of torture in sight of the host. At the convention Indutiomarus declared Cingetorix an enemy and confiscated his goods. […] This business despatched, Indutiomarus declared in the convention that he had been summoned by the Senones, the Carnutes, and several other Gallic states, and that he proposed to march to them through the borders of the Remi, laying waste their lands, and before so doing to attack the camp of Labienus. He gave instructions as to what he would have done” (Caesar, De Bello Gallico V, 56). Figure 6. Plan of the so-called Viereckschanze from Donnersberg. After Zeeb-Lanz (2008). Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 26 the Dutch river area—in more detail. Finally, we will discuss the strategic role of the regional cult places in the formation of politicized ethnic identities and the relationship with a specific deity worshipped in these sanctuaries. In the Early and Middle Iron Age (ca. 800–250 BC), we lack evidence in the Lower Rhine region for cult places that functioned at a regional level. Several scholars have suggested that at the local level the many cemeteries or urnfields also functioned as cult places in this period in that they formed focal points in local identity constructions where cult practices related to ancestor worship and death rituals were carried out (Gerritsen 2003, Roymans and Kortlang 1999). At this local level, there seems to have been an overlap between burial communities and cult communities. In addition to that, regional ethnic groupings may have existed, but these did not yet manifest themselves as politicized ethnic units with an institutionalized power structure. In the Late Iron Age (250–15 BC), important changes can be observed, one of them being the appearance of rectangular ditched and/or palisaded enclosures that can be interpreted as local open-air cult places. They are much too large (sides of over 20 m) to be interpreted as grave enclosures. Nevertheless, they may still have been related to ancestor worship and death rituals, since in most cases they are located in or near cremation cemeteries. There they often belong to the oldest structures, which suggests a link with the founding of new local communities. Examples have been excavated at Wijshagen-Plokrooi (B), Someren-Waterdael, Mierlo-Hout, and Lomm (N). We can observe the development of a new category of regional sanctuaries in the Late Iron Age, thus resulting in a hierarchical differentiation of cult place. This development points to an increased importance of power configurations and related institutions at higher levels of political organization, which probably corresponded with the emergence of politicized ethnic and sub-ethnic communities. The regional cult places are striking because of the deposition of larger numbers of metal objects (in particular fibulae, coins, and weapons) in combination with ample evidence for ritual feasting. The regional significance of the sanctuaries is further underlined by the fact that they became monumentalized in the late 1st century AD and developed into “grand sanctuaires Gallo-Romains” (Roymans 2004). Recent investigations of the monumental cult places of Empel, Kessel-Lith, and Elst in the Batavian river area have added substantially to our knowledge of this category of sanctuaries. Striking features are their pre-Roman origin, their strong masculine and military associations, and their link with the popular deity Hercules Magusanus (Roymans 2009a). It has been repeatedly suggested that this largescale public assembly of men at arms described by Caesar could have taken place in the public space within the oppidum of Titelberg, whose 10 ha would have more than enough space to hold a meeting of this kind. Although we can never be entirely certain about this specific point, we know for certain, thanks to this passage from De Bello Gallico, that large collective encounters of this kind existed amongst the Treveri which, together with the archaeological data obtained from the civitas itself and other parts of temperate Europe, helps to shed more light on the purpose of the public spaces documented within the oppida (Fernández-Götz 2012; Fichtl 2010, 2012b). Once again we find an extremely close connection between politics and religion, since we believe that calling the Treveran assembly was both a political act (preparation for war, humiliation of the principal rival for power, etc.) and also a religious one (ritual sacrifice of the last warrior to arrive). Public Spaces in a Decentralized Landscape? Sanctuaries in the Lower Rhine Region In publications on Late Iron Age temperate Europe, the Lower Rhine region is usually situated outside or in the northern periphery of the La Tène culture, and characterized as a zone inhabited by societies with a totally decentralized settlement system lacking oppida or other central places. An image is created of relatively static Iron Age societies where structural changes only started off after incorporation into the Roman empire. This image needs considerable modification now because of the results of recent settlement research and the study of mobile material culture (Gerritsen 2003, Gerritsen and Roymans 2006). The study of Late Iron Age coinage, weaponry, glass bracelets and brooches point to a process of intense latènization of material culture in this area from the 2nd century BC onwards (Roymans 2009b). A phenomenon that we will discuss more closely here and which is of crucial importance for our broader understanding of the social dynamics in the Lower Rhine region, is the development of supra-local cult places. As already pointed out, cult places are key-sites in the identity construction of groups at different social levels, varying from local communities to large tribal confederations. For archaeologists, the investigation of cult places offers an underexplored methodological instrument for getting grip at the social and political functioning of communities above the level of co-resident local groups. We will first make some general observations on the emergence of cult places in the Lower Rhine region, and then describe three regional sanctuaries—all situated in Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 27 (Fig. 8). A large ritual find complex consisting of weaponry, bronze vessels, large quantities of pottery, and human as well as animal bones, was found in an old Meuse bed. A few hundred meters from this ritual find complex, many architectural remains were found used as spolia in a late Roman fortification. They belong to a classisized Gallo-Roman temple that originally stood in the direct surroundings, but whose exact location is not known. The many items of Late La Tène and early-Roman military equipment recovered from the find complex points to the worship of a god connected with the domain of warfare. In the context of the Lower Rhine region, Hercules Magusanus is the most plausible candidate (Roymans 2009a). Although the coin spectrum starts relatively late (silver rainbow staters from the 50s BC), the brooch spectrum (with Nauheim fibulae and brooches of Middle La Tène type) suggests that Kessel/Lith began to emerge as a site of regional significance since LT D1. We can conclude that LT D was a period of major social transformations in the Lower Rhine region. The appearence of regional cult places points at a process in which ethnic groups increasingly manifested themselves as political entities, probably with the historically documented institutions like kingship, public assemblies, and senates (Fernández- Götz 2011). The above sanctuaries can be seen as focal points where origin myths were commemorated, assemblies held, political decisions taken, and where a collective identity was forged. According to Caesar’s information, the three cult places were situated in the territory of the Eburones who were largely annihilated by Caesar after their revolt of 54 BC. Between 50 and 15 BC, the Batavians in this region originated from a mixture of a Germanic immigrant group from the right bank of the Rhine and local indigenous groups, remnants of the Eburones (Roymans 2004). Recent archaeological investigations in the sanctuaries of Empel and Elst have produced extensive evidence for a practice of ritual feasting. The bone material is heavily dominated by young cattle, which were slaughtered and consumed at the cult sites (Derks et al. 2008:98–116, Seijnen 1994). This feasting in public cult places was an important means of social interaction in early Batavian society. Powerful networks were sustained by collective food and drink rituals, and they no doubt constituted a major means of defining membership of the Batavian community. An interesting aspect is that we know the deity that was worshipped in the sanctuaries mentioned above: Hercules Magusanus. This god can be considered a syncretism of an indigenous The cult place of Empel is located on the south bank of the river Meuse at the top of a sand-dune. It was excavated in the years 1989–1991 (Roymans and Derks 1994). Although the site was heavily damaged by a levelling of the sand-dune in the early 1950s, we could reconstruct the groundplan of a “classicized” Gallo-Roman temple lying within a rectangular temenos wall (Fig. 7). This temple complex was constructed ca. 100 AD, but in its earliest phases the cult place seems to have been an open-air sanctuary enclosed by wooden fences following the contours of the sand-dune. At the eastern flank, there were several ritual foci in the form of two rows of posts and a cluster of pits dating to the Late Iron Age and earliest Roman period, where concentrations of metal finds were encountered. The Late Iron Age material includes over 800 coins and over 100 other metal objects, mainly brooches and fragments of military equipment. The practice of depositing coins has a rather late start in Empel, in the 50s BC, with a great number of gold staters ascribed to the Eburones (Roymans and Aarts 2005). The fibulae (with early Nauheim brooches and brooches of Middle La Tène construction), however, indicate an earlier start of the cult place in LT D1. The study of animal bone material produced evidence for ritual feasting. The finds are characterized by a cattledominated bone spectrum, indicating that cattle had been butchered and consumed at the cult site. A bronze figurine of Hercules and a votive inscription show that the central deity worshipped at Empel was Hercules Magusanus. The cult place of Elst lies in the eastern part of the Batavian civitas. In 1947, the remains of a monumental Gallo-Roman temple were excavated below the Christian church in the center of the village. Recent new research produced a fragment of a bronze figurine of Hercules and a fragment of a votive altar dedicated to Hercules, making it plausible that here too the deity Hercules Magusanus was worshipped (Derks et al. 2008, Roymans 2009a). Hardly any coins or other metal objects were found during the 1947 excavations, but recently some Late Iron Age rainbow staters and a LT D2 brooch were collected using a metal detector. The lowest find-layers from the pre-temple phase consisted of a compact layer of animal bones of predominantly young cattle. Together this evidence suggests that Elst, like Empel, was a sanctuary of pre-Roman origin. A third regional cult center was identified recently at Kessel/Lith, located on the south bank of the Meuse at the ancient confluence with the river Waal (Roymans 2004:103–193). Here, a major settlement with a cult place was destroyed during large-scale sand and gravel extraction between 1950 and 1990 Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 28 deity or hero called Magusanus and the Roman Hercules with whom he shared several structural characteristics in the early-Roman period (Roymans 2009a). It was the principal male deity of the Batavians and probably also of other Lower Rhine groups. His cult was closely associated with the domain of warfare and therefore heavily maleoriented. He may also have played a role as protector of the youths and as a mythical ancestor of the Lower Rhine groups. In the Roman period, the votive inscriptions and votive materials show his popularity among soldiers and ex-soldiers]. Magusanus, because of his martial capacities, seems to have been the symbol of the Batavian warrior ideology and the hero of the Batavian juventus. His pastoral associations directly linked him to cattle breeding as being the major subsistence strategy of many Lower Rhine groups. The above cult places must have been important places of aggregation of groups. It is useful to distinguish here between public meetings held at fixed times according to a ritual calendar, and ad hoc meetings organized in situations of crisis. An example of regularly held calendar meetings are the public initiation rites of juvenile peers into the warrior class; in tribal societies with martial traditions this rite of passage was often the moment when men received their first arms. Tacitus explicitly refers to this ritual (Germania 13): “yet the custom is that no one takes arms until the tribe has endorsed his future competence: then in the assembly itself one of the chiefs or his father or his relatives equip the young man with shield and spear: this corresponds with them to the toga, and is youth’s first public distinction: hitherto he seems a member of the household, next a member of the tribe”. For the Lower Rhine Figure 7. Simplified plan of the sanctuary of Empel. (1) foundations of Roman walls, (2) robber trenches of roman walls, (3) reconstructed Roman walls, (4) unexcavated Roman walls, (5) picket fences, (6) reconstructed picket fences, (7) posthole, (8) Roman well, (9) medieval well, (10) pit, (11) pleistocene sand, and (12) clay soil. After Roymans and Derks (1994). Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 29 region, it seems plausible to locate such public meetings where the initiation rituals were held at the cult places of Hercules Magusanus, who was intensively involved with the domain of warfare and thus the political domain. An example of a public assembly at a sanctuary in a situation of crisis can be found in Tacitus’ account (Hist. 4.14) of the proclamation of the Batavian revolt in 69 AD. In a sacred forest (sacrum nemus), where the Batavian leaders had gathered for a nocturnal banquet, Civilis succeeded in convincing his fellow tribesman to revolt against Rome. It is very possible that one of the three Hercules sanctuaries described above, was the location where the Batavians proclaimed their revolt. The only information we have about the physical appearance of the cult place is that it must have been dominated by trees. From this point of view, it is interesting to consider the results of the palynological research carried out at Empel (Groenman-van Waateringe and Pals 1994). The vegetation at the sanctuary can be characterized as a forest dominated by oak trees, and this is the case in the pre-temple period as well as in the temple phase. Performing Identities: Polities as Symbolic Constructs Communities are ultimately symbolic constructs (Cohen 1985). Given the fact that public cult places often functioned as lieux de mémoire where foundation myths were reproduced through rituals, cult celebrations, etc., we can conclude that these sites played a vital role in the symbolic construction of ethnic communities in antiquity and in the creation of boundaries with outside groups (Derks and Roymans 2009, Gerritsen and Roymans 2006). In fact, in recent years various authors have pointed out that there may have been a close link between the appearance of large cult centers like Gournay-sur- Aronde, Ribemont-sur-Ancre, and Mirebeau, and the emergence of politicized ethnic identities such as pagi and civitates in pre-Roman Gaul (Fichtl 2007, Wells 2006), similar to that demonstrated by de Polignac (1995) for the origins of the Greek poleis. The importance of assemblies and religious ceremonies as arenas for collective negotiation is thus clear, and the public spaces/sanctuaries of the Treveran oppida, but also sites like Kessel/Lith or Empel, would reflect precisely this need for meeting Figure 8. Palaeogeographic reconstruction of the confluence of the rivers Meuse and Waal at Kessel in the Late Iron Age and Early Roman period. (A) late medieval river embankments, (b) (sub-) modern river forelands, (c) presumed river course, (d) major zone with ritual depositions, and (e) Late Iron Age/Early Roman settlement complex. After Roymans (2004). Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 30 places. Over and above the differences, all the public spaces/sanctuaries referred to above have in common that they were places where communities came together and took part in fundamental activities for the social and biological reproduction of groups. At least some of the inhabitants of rural settlements dotted around the territory would meet at these sites at particular times of the year, normally coinciding with religious festivals or in response to specific circumstances, such as declaring war, as happened in the case of the Treveran assembly mentioned above. In an eminently rural world frequently characterized by transport difficulties, people would take advantage of these multitudinous encounters to deal with religious, social, economic, and political matters at the same time (see also de Ligt and de Neeve 1988). The role of public sanctuaries as key locations for the creation of collective group identities is graphically illustrated in Tacitus’ account of the Germanic Suebi and their central cult place: “They describe the Semnones as the most ancient and best-born of the Suebi. This credibility of their antiquity is confirmed by religion. At fixed seasons all tribes of the same name and blood gather through their delegations at a certain forest […]. And after publicly offering up a human life, they celebrate the grim initiation of their barbarous worship” (Tacitus, Germania 39). Another example that can be relevant in this context is the celebration of the common ancestry myth of the Latins. As noted by Cornell (1997:9), “there can be little doubt about the antiquity of a Latin myth of common ancestry, and of its central element, the cult of Jupiter on the Alban Mount. The annual celebration of this cult, known as the Latiar or Feriae Latinae, was an assembly of the representatives of all the Latin communities […]. The ceremony was an expression of ethnic solidarity, and constituted an annual renewal of the ties of kinship that the Latins believed they shared. Participation in the cult was a definition of membership; the Latins were those peoples who received meat at the annual festival of the Latiar”. Although it has to remain hypothetical, the huge amount of animal bones documented in the public space of Titelberg (Méniel 2008) could perhaps be related to ceremonies similar to the one described for the Latins, where membership was defined by the consumption of meat. In any case, ritual feasting at the public cult places was surely an important means of social interaction. During politico-religious festivals, public meals would be held and generally preceded by sacrifices and libations, as testified by the enormous quantity of animal bones found at Gallic sites such as Titelberg, Fesques, or Acy-Romance, or the large number of wine amphorae documented at sites like Corent or Lyon (Poux 2004). Collective food and drink rituals sustained powerful networks, constituting a major mechanism of defining membership (Aranda Jiménez et al. 2011, Dietler and Hayden 2001). These celebrations constituted essential arenas for political action, representing privileged opportunities for establishing and reinforcing the social order. From the above, we can conclude that major gathering places were of key importance in fostering notions of collective identity and legitimizing hierarchical relationships through the ritual and political nature of the activities that took place there (Beck and Wiemer 2009, Fernández-Götz 2013). This merging of social domains can be expected to have been instrumental in the appearance and strengthening of political and ethnic identities. Regional and supra-regional sanctuaries constituted fixed points of reference in the landscape, acting as centers that structured collective identities. These sites served as focal points and assembly places; they generated, promoted, and reaffirmed the sense of community cohesion through the celebration of events such as assemblies, religious festivals, and sometimes also fairs. This aspect of Late Iron Age central places deserves more explicit investigation, regardless of whether they take the form of open or defended settlements or meeting places in forests. Literature Cited Aranda Jiménez, G., S. Montón Subías, and M. Sánchez Romero (Eds.) 2011. Guess Who’s Coming To Dinner: Feasting Rituals in the Prehistoric Societies of Europe and the Near East. Oxbow Books, Oxford, UK. Beck, H., and H.-U. Wiemer (Eds.). 2009. Feiern und Erinnern. Geschichtsbilder im Spiegel antiker Feste. Verlag Antike, Berlin, Germany. Brücken, G. 2008. Die archäologischen Untersuchungen auf dem Bleidenberg bei Oberfell an der Mosel, Kreis Mayen-Koblenz. Berichte zur Archäologie an Mittelrhein und Mosel 13:231–316. Brunaux, J.-L., P. Méniel, and F. Poplin. 1985. Gournay I: Les fouilles sur le sanctuaire et l’oppidum (1975–84). Revue Archéologique de Picardie nº sp., Amiens, France. Caesar. 2006. De Bello Gallico. H.J. Edwards (Ed.). LCL, London, UK. Roymans, N. and T. Derks (Eds.) 1994. De tempel van Empel. Een Herculesheiligdom in het woongebied van de Bataven. Graven naar het Brabantse verleden 2, `s- Hertogenbosch. Cohen, A.P. 1985. The Symbolic Construction of Community. Ellis Horwood, London, UK. Cornell, T. 1997. Ethnicity as a factor in early Roman history. Pp. 9–21, In T. Cornell and K. Lomas (Eds.). Gender and Ethnicity in Ancient Italy. Accordia Specialist Studies on Italy 6, London, UK. Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 31 Derks, T. 1998. Gods, Temples, and Ritual Practices. The Transformation of Religious Ideas and Values in Roman Gaul. Amsterdam University Press, Amsterdam, The Netherlands. Derks, T. and N. Roymans 2009. Introduction. Pp. 1–10, In T. Derks andN. Roymans (Eds.). Ethnic Constructs in Antiquity: The Role of Power and Tradition. Amsterdam University Press, Amsterdam, The Netherlands. Derks, T., J. van Kerckhove, and P. Hoff. 2008. Nieuw archeologisch onderzoek rond de grote kerk van Elst, gemeente Overbetuwe (2002–2003). Zuidnederlandse Archeologische Rapporten 31, Amsterdam, The Netherlands. Dietler, M., and B. Hayden (Eds.). 2001. Feasts: Archaeological and Ethnographic Perspectives on Food, Politics, and Power. Smithsonian Institution Press, Washington, DC, USA. Earle, T. 1997. How Chiefs Come to Power: The Political Economy in Prehistory. Stanford University Press, Stanford, CA, USA. Engels, H.-J. 1976. Der Donnersberg I. Die Viereckschanze. Franz Steiner Verlag, Wiesbaden, Germany. Fernández-Götz, M. 2011. Niveles sociopolíticos y órganos de gobierno en la Galia de finales de la Protohistoria. Habis 42:7–26. Fernández-Götz, M. 2012. Die Rolle der Heiligtümer bei der Konstruktion kollektiver Identitäten: das Beispiel der treverischen Oppida. Archäologisches Korrespondenzblatt 42(4):509–524. Fernández-Götz, M. 2013. Politik, Religion, und Jahrmärkte: Zur Rolle der Volksversammlungen im eisenzeitlichen und frühmittelalterlichen Europa. Pp. 71–82, In R. Karl and J. Leskovar (Eds.). Interpretierte Eisenzeiten 5. Fallstudien, Methoden, Theorie. Oberösterreichisches Landesmuseum, Linz, Austria. Fernández-Götz, M. 2014. Identity and Power: The Transformation of Iron Age Societies in Northeast Gaul. Amsterdam University Press, Amsterdam, The Netherlands. Fichtl, S. 2007. Le IIIe s. av. n. è.: genèse des entités politiques en Gaule? Pp. 283–289, In C. Mennessier- Jouannet, A.-M. Adam, and P.-Y. Milcent (Eds.). La Gaule dans son contexte européen aux IVe et IIIe siècles avant notre ère. Édition de l’Association pour le Développement de l’Archéologie en Languedoc- Roussillon, Lattes, France. Fichtl, S. 2010. Les places publiques dans les oppida. L’Archéologue, archéologie nouvelle 108:36–40. Fichtl, S. 2012a. Les peuples gaulois. IIIe-Ier siècle av. J.-C. Errance, Paris, France. Fichtl, S. 2012b. Places publiques et lieux de rassemblement à la fin de l’âge du Fer dans le monde celtique. Pp. 41–53, In A. Bouet (Ed.). Le forum en Gaule et dans les régions voisines. Ausonius, Bordeaux, France. Fleischer, F., and S. Rieckhoff 2002. Bibracte: Eine keltische Stadt. Das gallo-römische Oppidum auf dem Mont Beuvray (Frankreich). Pp. 103–118, In H.-U. Cain and S. Rieckhoff (Eds.). Fromm – Fremd – Barbarisch. Die Religion der Kelten. Verlag Philipp von Zabern, Mainz, Germany. Fritsch, T. 2009. Archäologische Forschungen am Hunnenring bei Otzenhausen. Pp. 123–129, In G. Uelsberg and M. Schmauder (Eds.). Kelten am Rhein. Akten des dreizehnten Internationalen Keltologiekongresses. Beihefte Bonner Jahrbücher 58/1, Bonn, Germany. Gerritsen, F. 2003. Local identities. Landscape, and Community in the Late Prehistoric Meuse-Demer-Scheldt Region. Amsterdam University Press, Amsterdam, The Netherlands. Gerritsen, F., and N. Roymans. 2006. Central places and the construction of tribal identities. The case of the Late Iron Age Lower Rhine region. Pp. 251–266, In C. Haselgrove (Ed.). Celtes et Gaulois, l’Archéologie face à l’Histoire. 4: Les mutations de la fin de l’âge du Fer. Actes de la table ronde de Cambridge, 7–8 juillet 2005. Collection Bibracte 12/4, Centre archéologique européen, Glux-en-Glenne, France. Groenman-van Waateringe, W., and J.-P. Pals 1994. De vegetatie op en rondom het heiligdom. Pp. 72–81, In N. Roymans and T. Derks (Eds.). De tempel van Empel. Een Herculesheiligdom in het woongebied van de Bataven. `s-Hertogenbosch, The Netherlands. Hornung, S. 2012. Le «Hunnenring » d’Otzenhausen, Lkr. St. Wendel, Sarre. Pp. 183–216, In M. Schönfelder and S. Sievers (Eds,), L’âge du Fer entre la Champagne et la vallée du Rhin. Verlag RGZM, Mainz, Germany. Kaczynski, B. 2009. Überlegungen zur Organisation des Münzwesens der Treverer. Pp. 199–204, In G. Uelsberg and M. Schmauder (Eds.), Kelten am Rhein. Akten des dreizehnten Internationalen Keltologiekongresses. Beihefte Bonner Jahrbücher 58/1, Bonn, Germany. Krausse, D. 2006 Eisenzeitlicher Kulturwandel und Romanisierung im Mosel-Eifel-Raum. Verlag Philipp von Zabern, Mainz, Germany. Krausse, D., and C. Nübold. 2007. Der Tempelbezirk von Wallendorf. Pp. 277–281, In G. Uelsberg and M. Schmauder (Eds.). Kelten am Rhein. Akten des dreizehnten Internationalen Keltologiekongresses. Beihefte Bonner Jahrbücher 58/1, Bonn, Germany. Ligt, L. de, and P.W. de Neeve 1988. Ancient periodic markets: Festivals and fairs. Athenaeum 66:391–416. Méniel, P. 2008. Les restes d’animaux de l’espace public de l’oppidum du Titelberg. Pp. 167–173, In D. Castella and M.-F. Meylan Krause (Eds.). Topographie sacrée et rituels. Le cas d’Aventicum, capitale des Helvètes. Archéologie suisse, Basel, Switzerland. Metzler, J. 1991. Les sanctuaires gaulois en territoire trévire. Pp. 28–41, In J.-L. Brunaux (Ed.). Les sanctuaires celtiques et leurs rapports avec le monde méditerranéen. Errance, Paris, France. Metzler, J. 1995. Das treverische Oppidum auf dem Titelberg (G.-H. Luxemburg). Zur Kontinuität zwischen der spätkeltischen und der frührömischen Zeit in Nord-Gallien. Dossiers d’Archéologie du Musée National d’Histoire et d’Art 3, Luxembourg. Metzler, J. 2006. Religion et politique. L’oppidum trévire du Titelberg. Pp. 191–202, In C. Goudineau (Ed.). Religion et société en Gaule, ed. C. Errance, Paris, France. Journal of the North Atlantic M. Fernández-Götz and N. Roymans 2015 Special Volume 8 32 Metzler, J. 2008. Du Titelberg à Trèves. De l’oppidum gaulois à la ville romaine. Pp. 155–165, In D. Castella and M.-F. Meylan Krause (Eds.). Topographie sacrée et rituels. Le cas d’Aventicum, capitale des Helvètes. Archéologie suisse, Basel, Switzerland. Metzler, J., and J. Scheid. 2011. Sanctuaires et organisation politique et sociale avant et après la conquête. L’exemple des grands lieux de culte de l’Est de la Gaule. Pp. 533–542, In M. Reddé et al. (Eds.). Aspects de la Romanisation dans l’Est de la Gaule. Collection Bibracte 21, Centre archéologique européen, Glux-en- Glenne, France. Metzler, J., P. Méniel, and C. Gaeng. 2006. Oppida et espaces publics. Pp. 201–224, In C. Haselgrove (Ed.).. Celtes et Gaulois, l’Archéologie face à l’Histoire. 4: Les mutations de la fin de l’âge du Fer. Actes de la table ronde de Cambridge, 7–8 juillet 2005. Collection Bibracte 12/4, Centre archéologique européen, Gluxen- Glenne, France. Nickel, C. 2013. Martberg – Heiligtum und Oppidum der Treverer. III: Die Siedlung. Funde und Befunde sowie naturwissenschaftliche Ergebnisse der Grabungen 1986/87 und 1994–2010. Berichte zur Archäologie an Mittelrhein und Mosel 19, Koblenz, Germany. Nickel, C., M. Thoma, and D. Wigg-Wolf. 2008. Martberg. Heiligtum und Oppidum der Treverer I. Der Kultbezirk. Die Grabungen 1994–2004. Berichte zur Archäologie an Mittelrhein und Mosel 14, Koblenz, Germany. Nortmann, H. 2009. Römisches Heiligtum und Theater in Kastel-Staadt. Jahrbuch Kreis Trier-Saarburg 2009:136–144. Nortmann, H., and A. Peiter. 2004. Kastel-Staadt: Ein Führer zu den archäologischen und historischen Zeugnissen. Rheinische Kunststätten 481, Neuss, Germany. Polignac, F. de 1995. Cults, Territory, and the Origins of the Greek City-State. The University of Chicago Press, Chicago, IL, USA. Poux, M. 2004. L’âge du vin. Rites de boisson, festins et libations en Gaule indépendante. Éditions Monique Mergoil, Montagnac, France. Roymans, N. 2004. Ethnic Identity and Imperial Power: The Batavians in the Early Roman Empire. Amsterdam University Press, Amsterdam, The Netherlands. Roymans, N. 2009a. Hercules and the construction of a Batavian identity in the context of the Roman empire. Pp. 219–238, In T. Derks and N. Roymans (Eds.). Ethnic Contructs in Antiquity: The Role of Power and Tradition. Amsterdam University Press, Amsterdam, The Netherlands. Roymans, N. 2009b. On the latènisation of Late Iron Age material culture in the Lower Rhine/Meuse area. Pp. 99–114, In J. Cession-Louppe (Ed.), Les Celtes aux racines de l’Europe. Monographies du Musée royal de Mariemont 18, Musée royal de Mariemont, Belgium. Roymans, N., and J. Aarts. 2005. Coins, soldiers and the Batavian Hercules cult. Coin deposition in the sanctuary of Empel in the Lower Rhine region. Pp. 337–359, In C. Haselgrove and D. Wigg (Eds.). Iron Age Coinage and Ritual Practices. Studien zu Fundmünzen der Antike 20, Mainz, Germany. Roymans, N., and T. Derks (Eds.) 1994. De tempel van Empel. Een Herculesheiligdom in het woongebied van de Bataven. Graven naar het Brabantse verleden 2, `s-Hertogenbosch. Roymans, N., and F. Kortlang. 1999. Urnfield symbolism, ancestors and the land in the Lower Rhine region. Pp. 33–61, In F. Theuws and N. Roymans (Eds.). Land and Ancestors: Cultural Dynamics in the Urnfield Period and the Middle Ages in the Southern Netherlands. Amsterdam University Press, Amsterdam, The Netherlands. Seijnen, M. 1994. Dierebotten en rituele maaltijden. Pp. 162–173, In N. Roymans and T. Derks (Eds.). De tempel van Empel. Een Herculesheiligdom in het woongebied van de Bataven. Graven naar het Brabantse verleden 2, `s-Hertogenbosch, The Netherlands. Tacitus. 1980. Germania. M. Hutton (Ed.)., rev. E.H. Warmington. LCL, London, UK. Tacitus. 1979–1980. Historiae. C.H. Moore (Ed.).LCL, London, UK. Thoma, M. 2000. Der gallo-römische Kultbezirk auf dem Martberg bei Pommern an der Mosel, Kr. Cochem- Zell. Pp. 447–483, In A. Haffner and S. Schnurbein (Eds.). Kelten, Germanen, Römer im Mittelgebirgsraum zwischen Luxemburg und Thüringen. Habelt, Bonn, Germany. Thoma, M. 2007. Vom Versammlungsplatz zum Heiligtum. Zur Genese einer gallorömischen Tempelanlage der Treverer a. d. Mosel (Deutschland). Pp. 55–76, In S. Groh and H. Sedlmayer (Eds.). Blut und Wein. Keltisch-römische Kultpraktiken. Éditions Monique Mergoil, Montagnac, France. Trunk, M. 2007. Anmerkungen zum römischen Theater von Belginum. Pp. 321–332, In R. Cordie (Ed.). Belginum. 50 Jahre Ausgrabungen und Forschungen. Verlag Philipp von Zabern, Mainz, Germany. Wells, P.S. 2006. Objects, meanings, and ritual in the emergence of the oppida. Pp. 139–153, In C. Haselgrove (Ed.), Celtes et Gaulois, l’Archéologie face à l’Histoire. 4: Les mutations de la fin de l’âge du Fer. Actes de la table ronde de Cambridge, 7–8 juillet 2005. Collection Bibracte 12/4, Centre archéologique européen, Glux-en-Glenne, France. Wenskus, R. 1961. Stammesbildung und Verfassung. Das Werden der frühmittelalterlichen gentes. Böhlau, Köln/Graz, Germany. Zeeb-Lanz, A. 2008. Der Donnersberg. Eine bedeutende spätkeltische Stadtanlage. Archäologische Denkmäler in der Pfalz 2, Speyer, Germany. Zeeb-Lanz, A. 2012. Neue Erkenntnisse zu Mauerarchitektur und Bauorganisation des keltischen Oppidums auf dem Donnersberg. Pp. 217–241, In M. Schönfelder and S. Sievers (Eds). L’âge du Fer entre la Champagne et la vallée du Rhin. Verlag RGZM, Mainz, Germany.