In chapter 5 I presented an historical model of the genesis of the Batavians in the Dutch river delta. My central hypothesis is that the formation of a Batavian identity group had its roots in the Caesarian frontier organisation: it emerged from a process of integration between a relatively small immigrant group from the east bank of the Middle Rhine and local indigenous groups. Tacitus describes the Batavians as a branch of the Chatti who had split off in order to settle in the Rhine delta. This move can be dated to the period between Caesar's departure from Gaul (51 BC) and the start of Drusus’ Germanic campaigns (12 BC). The objective of the present chapter is to test this historical model against a numismatic data set relating to the Lower Rhine ‘rainbow cups’ of the triquetrum type. The phase in which these coinages circulated widely – and were probably also struck – in the Rhine delta coincides with the historically documented formation of a Batavian polity. It will be argued that most of the coinages in question were Batavian emissions. Finally, I shall discuss the role these coinages may have played in the integration process of different groups into a new Batavian polity.

DISTRIBUTION, CLASSIFICATION AND CHRONOLOGY OF THE LOWER RHINE TRIQUETRUM COINAGES

In the Late Iron Age, the Lower Rhine region north of Bonn belonged to the northern periphery of coinusing communities. At first pre-Roman coin circulation was a marginal phenomenon. Not until the mid-1st century BC did coin usage expand considerably and local coin production develop. Apart from the more recent AVAVCIA coins of the Scheers 217 type, the so-called ‘rainbow cups’ of the triquetrum type were the principal local coin group. These feature on the convex obverse a cup-shaped profile with a triquetrum inside a ‘laurel wreath’, and on the concave reverse a ‘pyramid’ of point circles and double circles surrounded by a zigzag line. The triquetrum coins were minted in various metals: gold/electrum (rare in the Lower Rhine region), silver, and a copper alloy. Their weight fluctuates between 7.5 and 4.5 grams. The history of this coin group, which spans almost the entire 1st century BC, began in the area east of the Middle Rhine.

The primary aim of this study is to arrive at a model of Batavian ethnogenesis in the specific context of the Germanic frontier of the Roman empire. This implies both the reconstruction of ethnogenetic processes and their political context, and an attempt at reconstructing the image and self-image of the Batavian community. With this study I hope to make a contribution to the broader discussion of ethnicity and ethnogenesis in antiquity. My approach is a ‘historical-anthropological’ one, employing concepts and insights from both the social and the historical sciences, as well as a micro/macro-perspective that analyses local developments against a broader historical backdrop. I also look at comparative historical research. However, this comparison is not only driven solely by the desire to make generalisations but also by the search for a better understanding of historical specifics. My research focuses on the period from the 1st century BC to the Batavian revolt in AD 69/70, but this timeframe is broadened where relevant.

Why the Batavians? The relative wealth of both historical and archaeological evidence makes them an attractive field of study. Their reputation as troop suppliers to the Roman army and the occasion of the Batavian revolt of 69/70, described at length in the surviving books of Tacitus’ Historiae, has earned them considerable attention in the literary sources. Their role as a soldiering people has created a comparatively rich database of epigraphic material in the form of epitaphs, votive inscriptions and military diplomas. A great deal of archaeological research has been carried out at the civitas Batavorum, both into its urban and military centre at Nijmegen and the settlements and cemeteries in the countryside. And recently, fundamental advances have been made in numismatic research and the study of public cult places. All this material has enabled us to put together an interesting case study of the Batavians, one which is also relevant for the broader discussion on ethnic identities in the Roman empire. In the context of Dutch archaeology, this book introduces a new field of research that is relevant for the study of both civilian and military aspects of the Lower Rhine frontier zone.

INTRODUCTION

Central to this chapter are gold staters of the Scheers 31 type, with a triskeles or whorl on the obverse and a horse facing left on the reverse. This coinage is interesting for several reasons:

• 1. It represents the most northerly Late La Tène gold emission on the European continent. Pre-Roman coin circulation was a peripheral phenomenon in the Lower Rhine region, which raises the question as to what factors determined the slow acceptance of coins in this area.

• 2. The relatively late date and limited distribution of the coinage offers various possibilities for historical interpretation. Several scholars have ascribed it to the Eburones and proposed a direct link with the Eburonean revolt against Caesar in 54/53 BC. However, the question is whether this interpretation remains plausible following the recent revision of the chronological framework for Late Iron Age coinage in Belgic Gaul.

• 3. It may tell us about the politico-geographical situation in the Lower Rhine region, and the Rhine/ Meuse delta in particular, at the time of Caesar's conquest – that is, prior to the ethnogenesis of the Batavians. A gold coin emission is direct evidence of a political authority which used the coins to consolidate and enforce its power networks.

• 4. The use of metal detectors has led to an upsurge in single coin finds of the Scheers 31 type in the past decade, with a two-fold increase in the number of sites and a three-fold increase in the number of coins. As a result, we are now better informed about the metrology, distribution and archaeological contexts of the coins. The latter enables us to focus on the depositional processes by which the coins ended up in the soil.

Using the new data, I will attempt to sketch the production, circulation and deposition of the gold triskeles coins and to answer the questions raised.

LATE IRON AGE COIN CIRCULATION IN THE LOWER RHINE REGION

In the Late Iron Age, the Lower Rhine region north of Bonn represented the northern periphery of the coin-using communities. Originally, pre-Roman coin circulation was a marginal phenomenon in this region.

I emphasised in my opening chapter the need, when studying Batavian ethnogenesis, to distinguish between the formation of the Batavians as a socio-political entity and their genesis as an ethnic group. This latter topic will be discussed in chapters 10 and 11. Central to the present chapter is the formation of the Batavians as a political community. I rely for the most part on historical sources. My investigation centres on the period from the mid-1st century BC until the Augustan era. This will involve some anticipation of the discussion of the political organisation of the later, pre-Flavian civitas Batavorum, which is the subject of chapter 8.

A reasonable argument can be made on archaeological grounds for regarding the Batavians in the Lower Rhine area as a newly formed tribe. Although Tacitus does refer to them as a splinter group of the Germanic Chatti who settled in an uninhabited part of the Rhine delta, the archaeological evidence suggests that they developed from complex, multi-ethnic origins. They were not simply a group of newcomers, but – in view of the considerable local continuity of settlement and material culture – included the remnants of older indigenous (probably Eburonean) groups as well. There emerged a new tribal association, a process which must have occurred in the course of a single generation.

The following discussion will cover three topics: the earliest history of the Romano-Batavian alliance, the probable role of a client kingship in this formative phase, and the introduction of an aristocratic order with a magistrature.

THE ROOTS OF THE ALLIANCE BETWEEN THE ROMANS AND BATAVIANS

According to Tacitus, the Batavians had a special alliance (antiqua societas) with the Romans in pre-Flavian times. The treaty regulated the supply of auxiliary troops in closed ethnic units and under their own commanders, and exempted the Batavians from paying tribute. Although the exact date of the treaty and the historical context in which it was made are unknown, it is generally assumed to have already been in place in 12 BC when Drusus used the Batavian region as a base for his campaigns into Germania magna. Willems has argued that the settlement of the Chatto-Batavian immigrant group in the Rhine delta somewhere between 50 and 12 BC was not a spontaneous process but was linked to Roman frontier policy.

THE ETHNOGENESIS OF THE BATAVIANS. A SUMMARY

In this study I have attempted to sketch a picture of Batavian ethnogenesis in the context of the Roman frontier. My starting point was the current view in the social and historical sciences of ethnicity as a culturally determined, subjective construct that is shaped through interaction with an ethnic ‘other’. This study sought to analyse literary, epigraphic and archaeological sources relating to the Batavian image and self-image against the background of the specific integration of the Batavian community into the Roman empire. The study's main conclusion is that we can demonstrate that Rome exerted a profound influence on the formation of the Batavians both as a political entity and as an ethnic group with its own historically anchored self-image. This impact goes far beyond what has been assumed to date and shows that we should regard the Batavians to a large extent as a creation of the Roman frontier.

The Batavians emerged as a political entity when a group from east of the Rhine was granted permission – possibly by Caesar himself during the Civil War period – to settle in the Rhine/Meuse delta on Gaul's northern periphery. This was a reward for past – and above all future – military support, laid down in a treaty by Rome. A new polity subsequently formed in the delta when the immigrant group amalgamated with the indigenous, former Eburonean population. At the political heart of this polity lay the Batavian stirps regia, the ancestors of Julius Civilis mentioned later by Tacitus. We could perhaps think here in terms of a client kingship, familiar to us from other frontier polities. The new royal dynasty will have been recognised and supported by Rome, as attested to by the early bestowal of Roman citizenship on this family, making them part of the clientele of the Julian house.

Archaeology, and more particularly the study of local coin emissions, is able to shed some light on the socio-political networks behind the emergence of a Batavian polity. In addition, there is evidence to suggest that the political and religious heart of the earliest Batavian polity was the already established central place of Kessel/Lith. Rainbow staters of the triquetrum type, which seem to have played a key role in the political integration of the various groups into a new tribal association, were possibly minted there.

INTRODUCTION

This chapter discusses an important complex of dredge finds retrieved by dredging personnel and amateur archaeologists during large-scale sand and gravel extraction at Kessel/Lith in recent decades. As is often the case with dredge finds, we have scant information about the specific archaeological contexts and we know only a very small part of the find complex. As a result, the finds have received little attention in the literature to date. Nevertheless, I propose to discuss the Kessel/Lith finds at some length in this chapter because their quantity and richness lends them considerable scientific importance. The Kessel/Lith site compels us to reconsider the prevailing image of Late Iron Age societies and their material culture in the Lower Rhineland.

As already stated in chapter 2, reference texts on Northwestern Europe in the later Iron Age present a stereotypical geographical division into a northern and a southern world, with the boundary usually running through southeast England, northern France and central Germany. The northern world is characterised by somewhat egalitarian and static societies with a barely differentiated settlement pattern consisting solely of dispersed hamlets and farmsteads. The southern world comprises the more dynamic, hierarchical, and complex societies of Gaul and Central Europe, characterised by the presence of oppida. These are viewed as the central places of tribal groups and are often assigned proto-urban characteristics. This spatial division is further reinforced by links to an ethnic dichotomy. The northern world is described as ‘Germanic’ and the southern as ‘Celtic’.

Textbooks usually include the Lower Rhineland in the northern, ‘Germanic’ world. Provincial Roman archaeologists point out that the process of Roman urbanisation proceeded much more slowly in this region because of the complete absence of a tradition of native centre settlements. Nijmegen is considered the oldest central place in the Rhine delta, fully initiated by the Roman authorities – the army in particular – and thus constituting an implantation from outside by a superpower.

This stereotypical picture has been questioned in recent years. The image of egalitarian, barely differentiated societies in the Lower Rhineland was largely prompted by the absence of a tradition of depositing weapons and personal ornaments in graves. But there is evidence to suggest that the dynamic and internal differentiation in the settlement pattern in the region has been underestimated for the pre- Roman period.

INTRODUCTION

This chapter deals with the political, institutional and territorial structure of the pre-Flavian civitas Batavorum in its relation to the Roman empire. It is a subject worthy of attention because the development of a Batavian identity group cannot be understood without reference to the political context. It is by no means straightforward, however. Historians and archaeologists are deeply divided about how the Batavian civitas was organised in a political and institutional sense and how it fitted within the Roman system of government. The discussion centres around three concepts: frontier, municipalisation and provincialisation. Provincialisation refers to the region's integration into a formal provincial structure in accordance with the Roman model. Municipalisation denotes the introduction of a Roman system of civic administration in line with the civitas model, with codified laws, elected magistrates and public priesthoods, which gave the empire a fairly uniform foundation. In terms of their legal status within the empire, the various entities could be a municipium or colonia, but also a community with a peregrine status. Frontier means the zone to which Rome had territorial claims, but which lay outside the provincialised core of the empire and which was subject to military authority. We can distinguish two main sides in the debate. The first assumes an early municipalisation of the Batavian civitas in the Augustan or Tiberian period and is best expressed by Raepsaet-Charlier. The second – recently reformulated by Slofstra – assumes a late municipalisation in the Flavian era. Slofstra suggests that up until the time of the Batavian revolt the Rhine delta should be viewed as a frontier zone, characterised by a specific political and institutional order. Of course, this difference of opinion is mainly the result of the incomplete and at times contradictory nature of the historical sources. However, it is also due in part to different ideas about how Roman imperial power was organised in the northern frontier zones. In this chapter I will compare and evaluate both sides of the debate (the frontier model versus the ‘civic’ or municipal model). I will then present my own view, based on new empirical data (such as the results of recent excavations at Waldgirmes in Germany) and comparisons with the situation in other frontier regions of the empire.

In chapter 5 I argued on the basis of historical sources that the Batavian ethnogenesis was closely bound up with Caesarian frontier politics, and I proposed the following historical reconstruction. At the time of the Gallic and subsequent Civil Wars, a treaty existed between Caesar and the leader of a Chatti-dominated group of east-bank Germans, who – probably in the 40s BC - were allocated land in the Rhine/ Meuse delta. The new Batavian polity arose when the dominant core of migrants from across the Rhine merged with indigenous groups. From the outset, then, the Batavian community's existence was closely tied to an alliance with Rome. At the heart of this alliance lay the supply of auxiliary troops, including a cavalry detachment that operated as personal bodyguard to Caesar and the later Julian emperors. In return, the Batavians were exempt from paying tribute and were granted the right to command their own troops. Their leader was probably formally recognised as king by Caesar, rewarded with Roman citizenship and thus incorporated in the clientela of the Julian house. Under Emperor Augustus, the old treaty with the Batavians was transformed into a public alliance with the Roman state.470 Although the kingship was abolished, the Batavian stirps regia retained its dominant political position. Until the death of Nero, the last of the Julio-Claudian line, there had been a solid alliance between Batavians and the emperor, which essentially built on the original treaty provisions.

Links with Rome from the Augustan period onward will have given rise to new types of self-representation among the Batavian community, and their elite in particular. In the Gallic provinces from that time we witness the adoption of new media like public writing and statuary to emphasise the links with the emperor and the imperial house. These new cultural forms also gave voice to local identities in the new context of the Roman empire. I am thinking in particular of the appearance among the Batavians of public monuments that symbolised the allegiance with Rome and the close ties with the Julio-Claudian house. An analogy with the Lingones in Gaul is instructive here. Julius Sabinus, a representative of an aristocratic family that was granted Roman citizenship early on, claimed direct descent from the deified Caesar, who allegedly had had a love affair with his great grandmother during Caesar's stay in Gaul.

INTRODUCTION

In recent decades the study of Late Iron Age societies in Gaul and the Rhineland has been at the forefront of discussion in both academic and popular archaeology. The primary focus has been the major social changes that occurred during that period, leading to more complex societies with a more highly developed social hierarchy and the first moves toward urbanisation. The most notable archaeological evidence is the appearance of major fortified settlements or oppida, a rapid rise in the use of coins and the emergence of collective sanctuaries. Such changes are usually regarded as diagnostic of the La Tène cultural region, distinguishing it from regions to the north where they did not occur.

In the Northwest European context, what picture do archaeological texts of today paint of Late Iron Age societies in the Lower Rhine region? The first response is that this region barely rates a mention in the international literature. It is viewed as part of the northern border zone of Gaul, as a region weak in La Tène cultural influences and which saw no structural social change in the Iron Age. Many publications reproduce a map showing the distribution of Late Iron Age oppida (fig. 2.1); the northern border runs through Northern France, Southern Belgium and the German Middle Rhine region towards Central Europe. The Lower Rhine region is thus usually seen as part of the northern zone of rather static societies with relatively egalitarian social structures. This picture stems chiefly from the Lower Rhineland's ‘poor’ material culture, in particular the weak presence of elements associated with elite power, such as major fortified settlements and rich metalwork. Characteristic are the simple burial ritual and the barely differentiated settlement pattern with an absence of oppida. Major social change, the texts suggest, would not occur there until after the Roman conquest. Some authors explain these regional differences by using core-periphery models or social evolutionary perspectives. There is also a long and powerful tradition of explaining the differences in ethnic terms (fig. 2.1). The Lower Rhine region is regarded as part of the northern ‘Germanic’ world, the counterpart of a southern ‘Celtic’ world.

INTRODUCTION

In chapter one of this book I outlined the essentials of ethnicity, defining ethnic identity as the temporary resultant of a process of developing collective self-images, attitudes and conduct that takes place in a context of interaction between those directly involved and outsiders. Ethnic identities are by definition subjective, dynamic and situational constructs, which makes their relationship to material culture problematic. Unlike many other kinds of cultural identity, unless combined with textual data, they are in principle archaeologically intangible. The purpose of this chapter is to elaborate on and apply these general principles in the specific case of the Batavians.

In chapter 5 I outlined a model of an emerging Batavian polity in the Lower Rhine frontier zone of the Roman empire and of the earliest beginnings of the Roman-Batavian relationship. It is against that historical background that I intend to analyse in this chapter the creation of a Batavian self-image, one which seems to have been shaped to a significant degree by interaction with the Roman world. It is clear that the Batavian community as a political entity came first, and that it was then followed by the shaping of an ethnic identity – a Batavian self-image.

By definition, a tension exists between ethnic identity as image or representation and as reality. As a rule, ethnic identities are constructed around a body of clichés, stereotypes and invented histories. A process is involved, whereby the collective formulates and applies rules of inclusion, role fulfilment and exclusion in interaction with their self-image and the image that others construct of them. But ethnic identity covers actions as well as images. It would be correct to say that ethnic identities are moulded, channelled and modified through constant interaction between the group image and the praxis of individual and collective action. In terms of the topic of my research, this means that Batavian identity was shaped in the forcefield between internal and external perception – between self-image and the image formed by outsiders – and was then named and appropriated as their own.

One obvious methodological problem is the almost total absence of primary sources about the selfimage of the Batavians.

The research programme entitled The Batavians. Ethnic identity in a frontier situation was launched at the Archaeological Centre of the Free University (ACVU) in Amsterdam in 1999. Supported by the Netherlands Organisation for Scientific Research (NWO), it is scheduled for completion in 2004. The study before you is part of this research programme and aims to provide a synthesis of the formation and earliest beginnings of the Batavian identity group in the context of the Roman empire.

I have worked on this book with considerable pleasure over the past few years, and am fortunate to have been able to do so in such a stimulating and supportive environment. I therefore wish to thank all my colleagues at the ACVU for their discussions on the many topics touched on in this book and for their critical comments and advice on draft versions of the individual chapters. In particular I would like to thank Joris Aarts, Ton Derks, Fokke Gerritsen, Jan Kolen, Johan Nicolay, Jan Slofstra, and Ivo Vossen. I highly appreciated the enthusiasm and encouragement which Jan Slofstra has always shown toward my research; I am glad that it was possible for us to work together for several years as colleagues at the same institute. For the illustrations and the final layout of the book I am indebted to Bert Brouwenstijn, who was sometimes assisted by Jaap Fokkema.

I would also like to express my appreciation to all the people and institutions that have helped in some way to make the publication of this book possible. I am grateful to Louis Swinkels and Annelies Koster of the Museum het Valkhof in Nijmegen for making photographic material available. Jan Thijssen and Harry van Enckevort provided information about some unpublished finds from Nijmegen. I wish to thank the State Archaeological Service (ROB) in Amersfoort for their warm hospitality during my sabbatical leave in 2003. And I would also like to thank the ROB, and W.J.H. Verwers in particular, for granting permission to publish the architectural remains from Kessel. I am grateful to Kees Peeterse (Pansa BV, Nijmegen) for his suggestions about the reconstruction of the temple of Kessel. Curators of various museums have helped me in my examination of the find material, in particular the Rijksmuseum van Oudheden in Leiden, the Noordbrabants Museum in ‘s-Hertogenbosch, Museum het Valkhof in Nijmegen, and the Galloromeins Museum in Tongres.

The castle was no less prominent in peace than war, above all, of course, because it was the residence of the great, the centre and the seat of lordship, and this in an age of lordship when a ruling class really ruled. From first to last, as we have seen, the castle was a fortified residence: the residential function was no less fundamental to it than the military; and it was, indeed, this unique duality of residence and fortress that, so to speak, made a castle, and made it different from the fortifications of earlier and later periods. One may say also that it made it feudal, for while it is a matter of historical fact that the castle, the fortified residence of a lord, is the peculiar manifestation of feudal society, it is also entirely appropriate that it should be so. Feudal society, we are told, is society organized for war. It is also most certainly a society dominated by a military and a militant aristocracy, and what more appropriate setting could there be for them than castles? That the seigneurial residence should also be a fortress fits perfectly, and makes manifest, the military ethos of the age, as also do the seals whereon these aristocratic warriors formally represented themselves, or the effigies they had placed upon their tombs, in both cases armed cap à pie. Of course not every lord in the feudal period lived in a castle all the time, and not all lordly residences were fortified, i.e. were castles; but also there is no doubt that the castle became the symbol as well as much of the substance of lordship, and thereby an architectural concept meant to impress. Those castles depicted as rising on the skyline of the Très Riches Heures are real. And meanwhile, also, the castle as the residence of the lord (or his official) became inevitably the centre of local government, as we shall see, and sometimes other things as well, arising from its strength and social eminence.

There is one general feature of this medieval high society which must be noticed before any further discussion of the castle as its characteristic residence. Its members were almost continually on the move.

The military rôle of the castle is the most obvious, the most romantic, and basically the most important. Though the castle was always a residence no less than a fortress, and though from these two fundamental rôles others subsidiary followed, it was military necessity which first called the castle into being, whether at the time of its origin in ninth- or tenth-century France or whether in the England of the Norman Conquest, and military necessity which caused precisely that fusion of the lordly residence and the stronghold which is the peculiar characteristic of the castle. It is, after all, the degree of fortification which distinguishes a castle from a house. Warfare in the earlier centuries at least turned first and foremost upon the castle, and though from the later fourteenth century the military importance of the castle may have begun to decline, to read of wars in the chronicles of the eleventh, twelfth and thirteenth centuries is largely to read of sieges, while the surviving records of English royal government, for example, show beyond doubt that the maintenance of castles and their fortification and preparation for war were primary concerns of contemporary military organization. If we begin to ask why this was so, one fundamental answer – by no means widely understood since medieval warfare is a widely neglected subject – is that the military rôle of the castle was not just defensive but also offensive. Indeed we may argue that the latter is primary, for it was the offensive capacity of the castle, its function as a base, heavily defended, for active operations by means of which the surrounding countryside could be controlled, that gave it much of its value in war, made it the prized object of attack, and thus accounts for all those sieges. Only in this way, therefore, is the defensive role of the castle its most characteristic, though certainly it is the sieges which attracted the limelight of recorded events. Because the base and residence should be as impregnable as possible, it is defence also which, above all other considerations and requirements, dictated the castle's design and architectural form in the centuries of its supremacy in war, even though it had to fulfil as well its other functions as lordly and prestigious dwelling, centre of local government and administration, and, it might be, treasury, armoury or prison.