Roman Imperialism and the Transformation of Rural Society in a Frontier Province: Diversifying the Narrative

Abstract

This paper studies rural populations in the Roman frontier province of Germania inferior, employing a perspective that allocates more space to the exploitative and repressive aspects of Roman rule. We draw attention to an alternative series of topics than the ones currently presented in rural archaeology. This includes attention to situations of crisis and instability, to fundamental reordering of rural populations, to issues of migration and to the interconnectivity of rural developments and imperial power structures. While these topics are usually considered as ‘historically given’, they are rarely the subject of serious archaeological research. This attempt at a more historicising approach does not mean a simple return to the traditional paradigm of historische Altertumskunde . Much better equipped than our predecessors of two or three generations ago, we archaeologists of the 21st century are able to engage in a critical and creative dialogue with historical sources and models.

Full text

Roman Imperialism and the
Transformation of Rural Society in
a Frontier Province: Diversifying
the Narrative
By NICO ROYMANS, TON DERKS and STIJN HEEREN
ABSTRACT
This paper studies rural populations in the Roman frontier province of Germania inferior,
employing a perspective that allocates more space to the exploitative and repressive aspects of
Roman rule. We draw attention to an alternative series of topics than the ones currently
presented in rural archaeology. This includes attention to situations of crisis and instability, to
fundamental reordering of rural populations, to issues of migration and to the interconnectivity
of rural developments and imperial power structures. While these topics are usually considered
as ‘historically given’, they are rarely the subject of serious archaeological research. This
attempt at a more historicising approach does not mean a simple return to the traditional
paradigm of historische Altertumskunde. Much better equipped than our predecessors of two
or three generations ago, we archaeologists of the 21st century are able to engage in a critical
and creative dialogue with historical sources and models.
Keywords: Roman imperialism; rural transformations; migration; reordering of populations;
historicising approach
INTRODUCTION: THEORETICAL AND METHODOLOGICAL CONSIDERATIONS
Over the course of the past three decades the study of rural populations in the northern
provinces of the Roman Empire has experienced a spectacular boost as a result of the
upsurge in new excavations generated by the introduction of development-led
archaeology. The enormous increase in archaeological data from the 1990s onwards has
stimulated new synthesising research on Roman rural settlement and land use. In the
Netherlands, research programmes led by the present authors have focused on the study of villa
landscapes, the transformation of Batavian society, the integration of peripheral rural
communities into the Roman world and the social dynamics on the late Roman frontier.
1
In
1
Roymans 2004; Roymans and Derks 2011; Roymans et al.2015;2017.
Britannia 2020, page 1 of 30
doi:10.1017/S0068113X20000148
© The Authors, 2020. Published by Cambridge University Press on behalf of The Society for the Promotion of Roman Studies. This is an
Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike licence
(http://creativecommons.org/licenses/by-nc-sa/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium,
provided the same Creative Commons licence is included and the original work is properly cited. The written permission of Cambridge
University Press must be obtained for commercial re-use.

Britain, the three-part New Visions of the Countryside of Roman Britain, edited by Michael Fulford
and Neil Holbrook, is a major achievement. The volumes on The Rural Settlement of Roman
Britain (Smith et al.2016), The Rural Economy of Roman Britain (Allen et al.2017) and Life
and Death in the Countryside of Roman Britain (Smith et al.2018) offer excellent regional
overviews; they are richly illustrated with high-level quantification and GIS presentations of the
primary data.
2
In France, the recently completed RurLand project directed by Michel Reddé
marks a milestone. The edited volumes Gallia Rustica I (2017) and II (2018) present a series
of regional and thematic studies covering northern and northeastern France, Belgium, the
southern Netherlands and the German Rhineland.
3
The papers offer a richly documented
overview of recent French research.
These and other studies employ different approaches to studying Roman period rural society.
By far the most attention is usually directed towards the description and analysis of the primary
evidence regarding house architecture, mobile material culture, burial practices, agrarian
strategies and regional settlement and land-use patterns. Many studies emphasise the remarkable
regional diversity in rural landscapes. This heterogeneity is seen as the result of contingent
factors working together, such as the varying natural potentials of landscapes for subsistence
strategies, the agency of Roman authorities in the sphere of taxation and exploitation of local
resources, and –especially in the cultural sphere –the relative autonomy of individuals and
groups in expressing their own identity in the context of the Roman world, thereby creatively
appropriating new ideas, lifestyles and material culture. Within this broad spectrum of
perspectives Reddé opts for a strongly economic approach, concentrating on the comparative
analysis and diachronic development of regional economies and settlement patterns.
4
In the past
two decades, British archaeology –influenced by the post-processual research agenda –has
profiled itself with studies on the cultural dimensions of ‘becoming Roman’and related identity
constructions, thereby often neglecting relations of inequality and interdependency between the
agents involved. David Mattingly, however, using a post-colonial approach, emphasises that
rural communities should always be studied in relation to hard imperial power structures and
that attention should be paid to the varying and often ambivalent attitudes of individuals and
groups towards Rome.
5
In this contribution we would like –much in line with Mattingly –to argue for a perspective on
rural societies that allocates more space to the exploitative and repressive aspects of Roman rule.
Particularly in the domain of rural archaeology, this perspective is often underdeveloped.
Dominant in Roman rural archaeology is the narrative of ‘romanisation’that often implicitly
reproduces a classical humanistic ideal of civilisation. Even though rural groups may have lived
far removed from the urban and military centres, we should realise that they were not situated
in a power vacuum. Their functioning and identity constructions should always be understood
in their relation to imperial power networks. This is a line of research that is deeply rooted in
Dutch research traditions, strongly determined by the geographic situation of the Netherlands in
the Germanic frontier zone of the Roman Empire.
6
From this frontier perspective we draw
attention to an alternative series of ‘big issues’with the aim of producing narratives other than
those currently presented in rural archaeology. While these topics are usually considered as
‘historically given’, they have rarely been the subject of serious archaeological research. The
perspective employed here is distinguished by its explicit attention to:
2
Smith et al.2016;2018; Allen et al.2017.
3
Reddé 2017;2018a.
4
cf. also Reddé 2018b.
5
Mattingly 2006, 353–63; Mattingly 2011. On the post-colonial perspective, see also Webster and Cooper 1996;
Huskinson 2000; Hingley 2005; Gardner 2013.
6
Bloemers 1978; Willems 1984; Slofstra 2002; Van Driel-Murray 2003; Roymans 2004. On the study of Roman
frontiers, see Whittaker 1994;2009; Breeze 2018.
NICO ROYMANS ET AL.2

FIG. 1. Tribal map of the Lower Rhine frontier zone in the Caesarian (above) and early Imperial (below) periods. The
population movements shown in the latter map are based on written evidence.
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 3

1. phases of fundamental reordering of rural populations;
2. issues of human mobility and migration;
3. understanding rural developments within broader historical contexts;
4. the interconnectivity of rural developments and imperial power structures.
This attempt at a more historicising approach does not mean a simple return to the traditional
paradigm of the historische Altertumskunde, where archaeology functioned as a historical
sub-discipline. We now have access to an advanced set of social and applied theories, a wide
range of methodologies, including new natural-science methods, a hugely augmented
archaeological dataset and –last but not least –a much improved chronological resolution of
the material evidence. Being much better equipped than our predecessors of two or three
generations ago, we archaeologists of the 21st century are able to engage in a critical and
creative dialogue with historical sources and models.
The aim of this paper is to present a regional case study in which we try to operationalise the
perspective outlined above. Our study area roughly corresponds to the territory of the military
district and later province of Germania inferior (FIGS 1and 5). Five themes will be discussed in
chronological order, all referring to major transformations of rural society and closely related to
changing imperial power relations. We are repeatedly confronted with human mobility and
migration, the genesis and dissolution of ethnic groups and situations of crisis and instability.
We start our discussion of each theme with a brief sketch of the historical framework, after
which we see what archaeology can contribute to the debate. We end with some concluding
remarks and prospects for further research.
CONQUEST AND THE ARCHAEOLOGY OF MASS VIOLENCE AND GENOCIDE
In his De Bello Gallico Julius Caesar gives extensive accounts of his military campaigns in the
northern frontier zone of Gaul. The tribes living in this zone, called Germani or Germani
cisrhenani, faced extreme forms of Roman mass violence and even genocide. Absolute low points
were the mass enslavement of the Aduatuci in 57 B.C., the annihilation of the Tencteri and
Usipetes in 55 B.C. and the scorched-earth campaigns against the Eburones of 53 and 51 B.C.
7
However, how seriously can we take Caesar’s reports about his use of mass violence? Opinions
among modern scholars vary greatly: some consider his descriptions as reflecting a historic reality,
while others argue that they are heavily biased by imperialist rhetoric and self-glorification, and
that the Roman conquest had relatively little impact on the local population.
8
A major problem is the archaeological invisibility of Caesar’s conquests in the Lower Germanic
frontier zone. A mobile army using temporary marching camps and deploying scorched-earth
strategies in an attempt to fight decentralised enemy groups leaves very few archaeological
traces. However, this situation has changed in the past decade. A battle-related find complex at
Kessel near the Meuse/Waal confluence can be interpreted as the probable site where Caesar
massacred the Germanic Tencteri and Usipetes.
9
Given the presence of concentrations of
Roman lead sling shot and several gold hoards deposited in the early 50s B.C., a late Iron Age
fortification at Thuin in central Belgium may be regarded as the probable oppidum of the
Aduatuci, where Caesar enslaved 53,000 people.
10
7
Caes., BGall. 2.29–35 (Aduatuci), 4.4–15 (Tencteri and Usipetes), 6.34, 6.44, 8.24–5 (Eburones).
8
For opposing opinions among German scholars, cf. Galsterer 1990, 117; Eck 2004,41–2.
9
Roymans 2018.
10
Roymans and Scheers 2012. Archaeological fieldwork is currently being conducted at Thuin by the Université
Libre de Bruxelles with the aim of testing further this hypothesis.
NICO ROYMANS ET AL.4

Furthermore, the archaeology offers the possibility to study the demographic effects of Caesar’s
conquests.
11
Starting from the assumption that his campaigns –if taken seriously –must have
caused a significant demographic regression in our study area, research has been conducted in
five test regions on the degree of settlement discontinuity at the transition from the late Iron
Age to the early Roman period. Proceeding from excavated and published settlements, all
investigated regions produced evidence for a substantial habitation discontinuity in the course
of the first century B.C.(FIG.2). Although other factors may also have been involved here and a
precise dating remains impossible, a connection with the Roman conquest is a very plausible
interpretation.
12
Also relevant is the investigation of the late Iron Age circulation of precious metal in the study
area. On the basis of coin-die analyses and cross-datings of recently discovered coin hoards of the
so-called Fraire/Amby horizon (FIG.3), we are obtaining a better picture of the circulation and
deposition of high-value coins in the mid-first century B.C.
13
The 50s B.C. show a clear peak in
coin hoarding, and this picture is reinforced by the many isolated finds of gold coins belonging
to this phase, which in many cases will have been deposited as ‘mini hoards’. This mid-first-
century hoarding peak can hardly be interpreted as a temporal upsurge in the accidental loss of
precious-metal coins; it must have been crisis related. A link with Caesar’s conquest seems
plausible.
14
In particular his campaigns against the Aduatuci and the Eburones will have
created large numbers of victims who could no longer recover their emergency concealments of
precious-metal coins.
We may conclude that the Lower Germanic frontier zone offers an interesting potential for
combined archaeological-historical research on the use of mass violence and genocide in
Roman imperial expansion. The 50s B.C. do indeed appear to have been a period of traumatic
experiences for the inhabitants of the northern periphery of Gaul. In line with Caesar’s reports,
large regions seem to have been transformed into landscapes of trauma and terror. The
discontinuity in the tribal maps from the Caesarian and early Imperial periods (FIG.1) also
suggests that we should take his reports about depopulation and mass violence seriously.
Finally, the question arises as to how one might explain the excessive Roman violence in this
frontier zone. Several factors may have played a role, such as the desire to take revenge for the
ambush of a Roman army by the Eburones and the absence of urbanised oppida that would
have been easy targets for the Roman army. However, it is also interesting to link the excessive
Roman violence to the extremely negative ethnic framing of Germani in Caesar’s war
account.
15
They are depicted as barbarians par excellence, as a population of bandits and
terrorists who were unsuitable for incorporation into the Empire.
16
This negative stereotyping of
Germani appears as a constant factor in Caesar’s narrative from his first war year in Gaul, and
it may have influenced his military actions against Germanic groups by removing barriers to
the use of extreme violence against them in conflict situations. In BGall. 6.9.7 we read that the
Germanic Ubii feared falling victim to ‘a general hatred of the Germans’(communi odio
Germanorum). This hatred may have reflected the general atmosphere in Caesar’s army.
11
For the methodology, primary evidence and results of this research, see Roymans 2019.
12
Of course, large-scale settlement discontinuity does not necessarily imply mass violence or massacres, but in
combination with historic evidence it may be a serious option.
13
Roymans and Scheers 2012, with interpretations based on the hoards of the so-called Fraire-Amby horizon.
14
cf. also the earlier views expressed in Haselgrove 1984; Nash 1987. For a critical view, see Haselgrove 2019,
where it is emphasised that the production of part of Simone Scheers’coinages ‘de type belge’had already started
before the 50s B.C.
15
BGall. 4.1–3, 6.21–8. See the discussion in Roymans 2019.
16
cf. Riggsby 2006,60–1; Schadee 2008; Krebs 2011;2018,93–122. On the imaginary geography of Caesar’s
Germania, see also Krebs 2006.
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 5

FIG. 2. Diagram of habitation trajectories of excavated rural settlements and cemeteries in the Meuse/Demer/Scheldt
area in the late Iron Age and early Roman period, showing a high level of habitation discontinuity in the later first
century B.C. and a large-scale recolonisation of the land in the Augustan period (after Roymans 2019, fig. 11).
NICO ROYMANS ET AL.6

Against this background it is perhaps no coincidence that four of the five cases of genocide
described in Caesar’s war account are clustered in the Germanic north (FIG.4).
THE AUGUSTAN REARRANGEMENT OF THE LOWER GERMANIC FRONTIER AND THE ARCHAEOLOGY OF
MIGRATION AND ETHNOGENESIS
Based on the written sources, the post-conquest period, and in particular the age of Augustus, can
be described as a formative phase for the Lower Germanic frontier zone, characterised by the
influx of new Germanic groups from the eastern bank of the Rhine, the formation of new tribes
and the first administrative ordering of the military district of Germania inferior. This resulted
over the course of the first century in a fundamental reorganisation of the tribal map compared
to that of the Caesarian period (FIG.1). Germania inferior most likely had six civitates (FIG.5):
the Ubii around Cologne, the Cugerni near Xanten, the Tungri with Tongres as their centre, the
Batavians around Nijmegen, the Cananefates with their capital at Voorburg and the Frisiavones
FIG. 3. Distribution of gold hoards of the Fraire-Amby horizon and hoards with silver triquetrum coins from the
mid-first century B.C.: (a) hoard with Gallo-Belgic gold staters; (b) hoard with silver rainbow staters; (c) mixed
hoard of Rhineland rainbow staters and Gallo-Belgic staters.
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 7

in the coastal area of Zeeland.
17
In addition, we hear about a number of smaller tribes, such as the
Texuandri, Baetasii and Sunuci, which were probably attributed (as pagi?) to one of the civitates
mentioned.
18
In several cases the scarce written sources inform us about the origin of the newcomers (FIG.1).
The Ubii were moved from the right to the left bank of the Rhine (the Cologne region) by Agrippa
in 39/38 or 20/19 B.C.
19
The Batavi –a subgroup of the Chatti who lived in modern Hesse –were
allowed to settle in an area said to be empty (vacua cultoribus) in the eastern half of the Dutch
Rhine delta at some time between Caesar’s departure from Gaul and the arrival of Drusus in
FIG. 4. Gaul at the time of the Roman conquest, with the distribution of the cases of genocide described by Caesar in
his Commentarii.
17
Raepsaet-Charlier 2002–03.
18
Texuandri: Roymans et al.2015. Baetasii and Sunuci: Rüger 1968, 97; Bechert 1982,55–6.
19
Strabo 4.3.4; cf. Tac., Ann. 12.27; Wolters 2001, 159–60.
NICO ROYMANS ET AL.8

about 15 B.C.
20
In 8 B.C. Tiberius transferred a group of 40,000 subject Sugambri and Suebi to the
Gallic bank of the Rhine, where they were probably settled in sparsely populated regions; this is
the only case for which we have information about the size of an immigrant population.
21
The
Cugerni and the Baetasii in the Xanten territory are often regarded as potential descendants of
this immigrant group.
22
The Cananefates in the Dutch coastal area had a close relationship with
the Chauci, and probably originated in the North Sea coastal area,
23
and the Frisiavones are
often considered the descendants of a group of Frisians transferred by the Roman general
Corbulo in 47.
24
Although this is not always explicitly mentioned, it is clear that the land
FIG. 5. Germania inferior and neighbouring provinces in the second and early third centuries (after Heeren 2017, fig. 1).
20
Tac., Germ. 29.1; Hist. 4.12; Roymans 2004,55–65.
21
Suet., Aug. 21; Tib. 9; Tac., Ann. 12.39; Wolters 2001, 163–7.
22
Rüger 1968,96–8; Bechert 1982,55–6; Heinrichs 2001. The Texuandri in the Meuse/Demer/Scheldt region are
also potential descendants of the Sugambri.
23
Tac., Ann. 11.18; De Bruin 2019, 7, 149–56.
24
Tac., Ann. 11.19; Raepsaet-Charlier 2002–03, 44; De Bruin 2019, 150.
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 9

allocations to Germanic groups were often directed or at least sanctioned by the Roman authorities,
thereby continuing a long tradition of rearranging both land and people in newly conquered
areas.
25
This settlement of new Germanic groups on the Gallic bank of the Lower Rhine meant a
fundamental change in Roman frontier policy from Caesar’s anti-Germanic policy based on
exclusion and negative ethnic framing. Under Augustus, Rome switched to a strategy of
incorporating Germanic groups on the right bank of the Rhine and transferring groups to the
heavily depopulated Gallic bank of the river. An important motif for this change of policy is
the repopulation of sparsely inhabited land and no doubt the exploitation of the military
potential of Germanic groups; from the Augustan period onwards groups in Germania inferior
were recruited on a massive scale for the military (see below).
What can archaeology add to the historical picture summarised above? While it is clear that
these changes must have had a profound impact on the rural populations that had been living
there, in archaeological terms almost no questions relating to this dynamic period have been
asked so far. To what extent can we see this influx of Germanic groups archaeologically? How
should we envisage this process of ethnogenesis of new groups? And what was the role of the
Roman authorities in all this? It is only recently that archaeology has become equipped with
the right methodological tools to study the theme of migration in the Lower Rhine frontier
zone. Over the past two decades significant progress has been made by viewing the new
Germanic frontier tribes first and foremost as political constructs arising from fusions of
autochthonous and immigrant groups under the direction of the Roman military authorities.
26
The archaeology of rural settlements can contribute to this debate by testing and improving the
essentially historic migration models. Important tools in this context are conventional material-
culture studies of personal ornaments, handmade pottery and indigenous house architecture, in
combination with strontium isotope studies of dental remains of domestic animals and chemical
analyses of pottery and metal ornaments. A specific pilot study is currently running in the
Batavian river area,
27
where we are trying to identify first-generation farmsteads of newly
founded rural settlements, thereby using the methods listed above to trace the geographic
origins of the settlers. The first results are promising and suggest that the influx of Germanic
groups was more diverse than the scarce written sources suggest. In the Batavian area, we
encounter a mixture of material culture of local origin, but also from the Rhine/Weser area and
the North Sea coastal region. On the basis of pottery style and house architecture, Jasper De
Bruin presumes that the settlers in the territory of the Cananefates originated from the North
Sea coastal area. German archaeologists point to the occurrence of ‘Elb-Germanic’domestic
pottery in rural settlements on the west bank of the Rhine in Ubian territory.
28
The most
plausible ethnogenetic model for the frontier tribes noted above is that they developed out of a
fusion of several immigrant groups of heterogeneous origin, possibly spread over time, that
were permitted or simply forced by the Roman military to settle down in exchange for
supplying auxiliaries. Larger incoming groups may have been granted the status of civitas,
25
cf. De Blois 2017; Woolf 2017. An interesting parallel for a forced deportation of a defeated people by Rome is
Livy’s account (40.38.2–3) of the fate of the Ligures in 180 B.C.: ‘The Liguri made frequent entreaties through
ambassadors that they might not be compelled to leave their household gods, the land in which they had been born,
the tombs of their ancestors, and also promised arms and hostages. When they met with no success and lacked the
strength to fight, they obeyed the edict.’The Liguri, about 40,000 free men with their wives and children, were
resettled on Roman state land, and both Roman consuls were awarded a triumph for their successful campaign.
26
cf. Roymans 2004 (Batavians); Raepsaet 2013 (Tungri); De Bruin 2019 (Cananefates).
27
Tiel-Medel as a Key Site for Innovative Research Towards Migration and Ethnogenesis in the Roman Frontier
(see Acknowledgements). On the methodology of migration studies of the Roman Empire, see Eckardt and Müldner
2016.
28
Andrikopoulou-Strack et al. 1999, 149–55; Hornung 2016, 297–302; Frank 2018; De Bruin 2019.
NICO ROYMANS ET AL.10

while smaller groups were usually allocated to an already existing civitas, where they may have
retained some degree of autonomy. The core of the immigrant groups usually consisted of a
war band that was obliged to offer its service to Rome.
We may conclude that the Augustan period was a phase of repopulation of the land and of the
formation of new ethnic groups that should be seen as political constructs which mask a much
more heterogeneous ethnic reality. The new ordering was to a large extent the product of
Roman frontier policy aimed at the large-scale exploitation of Germani as ethnic soldiers.
Archaeology can contribute to the discussion by testing and improving the current historical
models with regard to the migration and ethnogenesis of groups in the Lower Rhine frontier.
A BREEDING GROUND FOR SOLDIERS: THE IMPACT OF LARGE-SCALE AUXILIARY RECRUITMENT
A specific characteristic of the Lower Germanic region compared to many other frontier areas of the
Roman Empire was the high level of ethnic recruitment for the Roman army among indigenous rural
groups (FIG.6). This practice of ethnic soldiering was already in full swing in the Augustan-Tiberian
period and developed further over the course of the first century, when the initially irregular war
bands were transformed into regular formations of full-time soldiers. The impressive list of
pre-Flavian ethnic units suggests that the military government obliged each tribe to supply a
specific number of auxiliary units. This practice of ethnic recruitment created a close relationship
between native rural groups and the Roman military community. It confronts us with a specific
form of human mobility, one which primarily involved individuals who served in the army for a
long period and some of whom returned later as veterans. This latter practice in particular had an
enormous social impact in the first century since it brought inhabitants from almost every
settlement within a civitas into contact with the military variant of Roman culture.
This interaction between the rural and the military communities has been intensively explored
in Dutch archaeological research of the last two decades, which has relied heavily on the
systematic registration and analysis of metal-detection finds from private and public
collections.
29
Roman militaria are frequently encountered in rural contexts. For the Batavian
region a model has been proposed of the agency of returning auxiliary veterans in the spread of
the Latin language and script, including long-distance communication between soldiers and
their homeland, based on certain texts from the Vindolanda archive.
30
The close link with the
military domain brought about a rapid spread of Roman citizenship among the rural community
and the formation of military families that supplied soldiers generation after generation.
Certainly for the pre-Flavian period, it can be assumed that Roman pottery, coins, fibulae and
other ornaments mainly entered native settlements via military networks.
31
The intensive
appropriation of Roman mobile material culture went hand in hand with a conservative
adherence to the native byre-house traditions and burial practices. Wouter Vos links the
addition of a wooden portico to some indigenous farmhouses with the impact of veterans, who
were inspired by the military architecture of army camp barracks.
32
An indirect indication of
the crucial role of ethnic soldiering in Germania inferior, and in particular among the
Batavians, is the popularity of the martial Hercules cult in this region.
33
29
Nicolay 2007; Roymans 2011.
30
cf. Bowman and Thomas 1994–2003, 2.310, 2.346. Derks and Roymans (2002) interpret bronze seal-boxes from
rural settlements in the Rhine delta as indirect evidence for Latin literacy. Colin Andrews (2013) proposes an alternative
interpretation of the boxes as objects used for sealing packages of money. It is difficult to believe, however, that this
was the only way seal-boxes were used.
31
Roymans 2011; Roymans and Derks 2015.
32
Vos 2009, 237–51.
33
Roymans 2009.
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 11

This Batavian model cannot simply be extrapolated to the entire region of Germania inferior.
Nonetheless, almost all communities in the pre-Flavian period seem to have played a role as
providers of auxiliary units. The Batavians continued to cultivate their identity as a soldiering
people the longest; up into the third century, Batavian soldiers fostered their collective identity
in grave and votive inscriptions, in particular by using the tribal affiliation natione Batavus,
‘born a Batavian’.
34
In other districts this military tradition was abandoned much earlier,
especially in the emerging villa landscapes. Ethnic recruitment was most deeply rooted in the
poorer rural regions where there was an emphasis on animal husbandry. Military service
certainly offered opportunities to individuals as a source of income and a chance to acquire
citizenship and a legal marriage. However, the constant recruitment pressure will also have
raised tensions and given cause for excesses, as described by Tacitus in his account of the start
of the Batavian Revolt.
35
We see a different development in the civitates of the Ubii and the Cugerni, where coloniae
were founded in the towns of Cologne and Xanten, under the reign of Claudius and Trajan
respectively.
36
The founding of colonies implies the settlement of a group of legionary
FIG. 6. Overview of pre-Flavian ethnic recruitment by Rome in Germania inferior and Gallia Belgica (data after
Alföldy 1968): (A) civitates used for the conscription of auxiliary units; (B) ala;(C)cohors.
34
Roymans 2004, 253–8; Derks 2009.
35
Tac., Hist. 4.14.
36
Cologne: Eck 2004, 127–61; Lenz 2006. Xanten: Eck 2008.
NICO ROYMANS ET AL.12

veterans, which often meant a serious infringement on local landownership and power relations. It
is not clear, however, to what extent these veterans also settled in the countryside and acquired
land at the expense of the indigenous community. In the Cologne hinterland around Jülich
there is evidence for a systematic colonisation of previously uninhabited land in the Claudian
period.
37
The new settlers may have been retired legionaries or colonists from interior Gaul
but –given the layout of the farmsteads and the architecture of the earliest houses
38
–they
were not of local Germanic origin.
We may conclude that archaeology can contribute to the debate outlined above by enabling
study of the impact of ethnic recruitment, veteran behaviour and, more generally, the military
variant of Roman culture on rural communities. From a theoretical point of view it is
interesting that this impact occurred not only via a top-down élite model, but above all through
the agency of common soldiers. Finally, it is remarkable to observe that the short intermezzo of
the Batavian Revolt of A.D.69–70 appears to be almost untraceable in the rural archaeological
record; there are no indications of a break in the use of settlements and cemeteries, nor of a
demographic decline. The revolt certainly did not lead to a fundamental reordering of rural
habitation comparable to those attested there in other periods.
THE LATE THIRD-CENTURY CRISIS AND THE DEPOPULATION OF THE COUNTRYSIDE
For the later third century we are again confronted with a dramatic rural transformation in our
study area. According to the written sources, several interrelated political-military factors were
at play.
39
Wars on the frontiers of the eastern provinces as well as internal strife led to the
displacement of troops from the northern borders, which directly triggered Germanic invasions.
Many short-lived emperors and usurpers claimed the throne and lost their lives, often at the
hands of Roman rivals but some by ‘barbarian’armies. These upheavals were followed by
regional secessions, such as, in our area, the Gallic Empire (A.D. 259–74) founded by
Postumus. There are hints in the written sources that Postumus employed a double strategy
with regard to the Germanic threat: he fought them using military means, but also concluded
alliances and settled them within the province in order to include barbarians in his army.
40
What can archaeology contribute to this discussion? The key issue by which rural archaeology
can distinguish itself is the study of the depopulation of the countryside. The phenomenon of
massive depopulation has been a long-held assumption, but is becoming more and more an
empirical reality. Large numbers of excavated settlements form the empirical basis for this
research. In all areas of Germania inferior we observe settlements coming to an end, although
the intensity of this varies regionally (FIG.7). In the Pleistocene sandy landscapes of northern
Belgium and the southern Netherlands (MDS region), as well as the coastal area of the
Cananefates, the depopulation seems to have been almost complete (FIG.8).
41
Here we observe
the large-scale abandonment of indigenous settlements in the second half of the third century,
and evidence of habitation in the early and middle part of the fourth century is very scarce.
42
In the Batavian and Traianensian area, more single finds of coins and crossbow brooches from
37
Lenz 2006; Gaitzsch 2011.
38
Heimberg 2002–03,68–70.
39
Drinkwater 2008,30–50.
40
De Boone 1954,29–35, in detail; Todd 2008, 444, more broadly.
41
MDS region: Heeren 2015;2017. Cananefatian region: De Bruin 2019, 213–19. See also Van Enckevort et al.
2017.
42
This contrasts with sites north of the Rhine, where dendrochronology has identified early and mid-fourth-century
activity and where plentiful house plans and mobilia have been found. Cf. Van Es 1967; Taayke et al. 2012.
Dendrochronology: Erdrich 1998 cites several examples.
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 13

rural sites hint at some habitation in the fourth century. However, in this region too, almost all
excavated settlements and cemeteries were abandoned in the late third century.
43
Rural depopulation as described for the northern regions, which were essentially non-villa
landscapes, was less complete for the more southern villa landscapes in the Cologne hinterland
and the Belgian loess region around Tongres. Several survey studies are available for the
Cologne hinterland. An older but still useful study is that by Michael Gechter and Jürgen
Kunow,
44
who used pottery from fieldwalking campaigns and excavations as a proxy of the
habitation history of several subregions. Regions north of the road from Tongres to Cologne
show decreasing site numbers from the second to the third century and a complete absence of
sites in the fourth century. Regions south of the road show survival rates (the number of
fourth-century villas compared to early third-century ones) of 71 per cent, 52 per cent and 30
FIG. 7. Germania secunda and neighbouring provinces in the early fourth century (after Heeren 2017, fig. 3).
43
Xanten area: Brüggler et al. 2017, 30. Batavian region: Willems 1984, 142–61; Heeren 2009, 201–13;
Vos 2009, 216.
44
Gechter and Kunow 1986.
NICO ROYMANS ET AL.14

FIG. 8. Diagram of habitation trajectories of excavated rural settlements in the Meuse/Demer/Scheldt region in the later
Roman period, showing an almost complete depopulation in the later third century and partial resettlement in the late
fourth/early fifth century (after Heeren 2015, table 5, with additions). Thick horizontal line: habitation period with good
evidence; thin horizontal line: dating evidence uncertain; red cross: supposed fire catastrophe; vertical red line:
dendrochronological date of well.
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 15

per cent. For the Aldenhovener Platte, Karl-Heinz Lenz argues that the decline had already begun
in about the middle of the third century. The number of early fourth-century villas is almost half
that of early third-century sites.
45
This is confirmed by the evidence from excavated settlements,
showing that many of the villa sites in the Cologne hinterland were abandoned by the late third
century.
46
The same is true for the Nervian countryside, where there is even evidence for a
‘desertification’of certain subregions.
47
The archaeological evidence for depopulation is further substantiated by palynological research.
Pollen diagrams from the hinterland of Cologne and the river area in the north show a substantial
increase in tree pollen and a decrease in settlement indicators from the second half of the third
century onwards, suggesting that many settlements were abandoned and that much of the land
was waste ground.
48
The pollen diagram from Kleefsche Beek in the hinterland of Xanten
shows a drastic decline in settlement-indicating pollen from the second half of the third
century, implying a return to a forest landscape at a level not reached since Neolithic times.
49
The depopulation of the countryside sketched above was not an isolated phenomenon but was
paralleled by a serious urban decline and in some cases even collapse. Several civitas capitals north
of the Bavai-Cologne line seem to have been given up. The civilian centres of Voorburg/Forum
Hadriani and Nijmegen/Ulpia Noviomagus were abandoned more or less completely in the late
third century.
50
The new fortification erected at Nijmegen-Valkhof can best be interpreted as a
military site, an interpretation that is strengthened by the recently proposed identification of this
site as Castra Herculis, one of the military sites rebuilt by Julian II in 358.
51
At Xanten a new
defensive circuit was set around a much reduced core. Since Xanten was probably now called
Tricensimae, a reference to the Legion XXX, and the surrounding countryside appears to have
been largely uninhabited, it seems likely that Xanten had lost its function as a civitas capital
and now primarily served as a military base.
52
Tongres and Cologne seem to be the only
civitas capitals of Germania inferior that survived the late third-century crisis, albeit on a much
reduced scale.
53
We can seriously question whether the administrative infrastructure of the northern civitates
survived the end of the third century if there were no civil centres and hardly any rural
settlements. The indirect consequences must have been dramatic, ranging from a substantial loss
of the provincial tax revenues to the dissolution of the traditional ethnic identity groups which
had their basis in the countryside. The collapse of the four northern civitates and the survival of
Tongres and Cologne corresponds well to the situation described in the Notitia Galliarum,an
appendix to the Notitia Dignitatum.
54
This document was compiled in about 423 but is often
assumed to represent a fourth-century situation in most cases.
55
The Notitia Galliarum states that
Germania secunda had only two civitates: the Agrippinensian metropolis and the Tungrian civitas.
56
Archaeologists have been debating the causes of the depopulations in Germania inferior. It is
clear that we are dealing here with a complex combination of supra-regional, long-term
45
Lenz 1999.
46
Gaitzsch 2011, 291–2 (overview of dendrochronological datings of wells from villa settlements); Päffgen 2012;
Brüggler et al.2017, 30.
47
Clotuche et al. 2017, 181.
48
Willems 1984, 267, fig. 137; Kalis and Meurers-Balke 2007; Kalis et al. 2008.
49
Kalis et al.2008, 43, 36.
50
Voorburg: De Jonge et al. 2006; De Bruin 2019, 238. Nijmegen: Willems et al.2009; Van Enckevort and
Thijssen 2014.
51
Verhagen and Heeren 2016.
52
Bridger 2003; Otten and Ristow 2008.
53
Tongres: Vanderhoeven 2017. Cologne: Eck 2004, 586–692.
54
Seeck 1876 (1962).
55
Scharff 2005.
56
Seeck 1876 (1962), Not. Gall. 5.
NICO ROYMANS ET AL.16

economic and environmental factors, and factors working at a more regional level. The civil wars
between Roman contenders, as well as the various barbarian incursions mentioned in the written
sources, are obvious candidates, but the Plague of Cyprian dated to the 250s could also have
played a role. Modern authors have proposed climate change with colder summer temperatures
in the northwestern provinces from A.D. 250 onwards, resulting in rising groundwater tables and
partial flooding of coastal wetlands.
57
Large-scale soil degradation may also have been a
significant factor.
58
Rural settlement research in the southern Netherlands and northern Belgium
has provided hardly any evidence of the burning down of the youngest farmsteads. If that were
the case, the deepened byre sections present in most farmhouses of this period would have
preserved remains of burnt layers, but this is attested only three times (see FIG.8). What might
help to solve this question in the future is a narrower dating of the depopulation. For the
moment we must be content with assuming a combination of long-term phenomena, such as a
slow soil degradation and climate change, and incidental factors such as warfare, as being
responsible for the rural depopulations.
The explanations mentioned above may indeed have caused a serious population decline, but
they cannot account for the relatively swift and near total depopulation of regions. In this
context it is important also to consider the role of imperial power as an active agent in the
process of massive depopulation, thereby referring to evidence for forced deportation of
groups.
59
With regard to the episode of the Gallic Empire and its aftermath, some clues can be
found in the written sources. Postumus, the founder of the Gallic Empire, is said to have
employed troops of Germanic descent in his war against Gallienus, which is supported by the
distribution of gold coins of Postumus in Germania magna.
60
Aurelian ended the Gallic
secession and brought the territories back under Roman rule (274), but his murder led to
further chaos. Two passages from a few years later point to the forced deportation of people
from the northern civitates. In the first, Maximian Augustus (286–305) is said to have settled
Franks in Nervian and Treveran wastelands, ‘by which they regained their old position and
were part of the law again’, thereby suggesting that these Franks had inhabited Roman land
earlier, no doubt in Germania inferior.
61
In the second, Constantius Chlorus is said to have
defeated, in around A.D. 293 or 297, a Frankish group that had settled earlier in Batavia, where
they had been commanded by a former native of the place, probably of Roman provincial
origin.
62
Constantius deported the defeated Franks to interior Gaul, forcing them ‘to lay aside
their weapons and fierceness’. Although some translators postulate this leader to be Carausius
the Menapian,
63
both Willem Jan De Boone and Willem Willems strongly argue that he should
be identified as Postumus.
64
57
Climate change: Haas 2006; McCormick et al.2012; Gouw-Bouman et al. 2019, who reconstruct a drop in
summer temperature of 1.5 degrees from A.D. 250 onwards; Riechelmann and Gouw-Bouman 2019. Flooding of
Dutch coastal wetlands: De Bruin 2019, 218–19.
58
Groenman-Van Waateringe 1983.
59
Heeren 2015 for the study area. On the forced deportation of groups in the Roman Empire, see also De Blois
2017; Woolf 2017,38–9.
60
SHA, Tyr. Trig. 6.2: et cum multis auxiliis Postumus iuvaretur Celticis atque Francicis in bellum cum Victorino
processit. Cf. De Boone 1954, 36; Willems 1984, 249–50. For the distribution of aurei of Postumus in Germania
Magna, cf. Schulte 1983; Degler 2017, map 1.
61
Pan. Lat. 8 (Nixon and Rodgers 1994), Constantio Caesari 21: tuo, Maximiane Auguste, nutu Nerviorum et
Trevirorum arva iacentia postliminio restitutus et receptus in leges Francus excoluit. Cf. De Boone 1954, 57.
62
Pan. Lat. 6 (Nixon and Rodgers 1994), Constantino Augusto 5.3: terram Bataviam, sub ipse quondam alumno
suo a diversis Francorum gentibus occupatam omni hoste purgavit. Cf. De Boone 1954,57–8.
63
Nixon and Rodgers 1994, 224, following La Baume.
64
De Boone 1954, 36, 42, 58; Willems 1984, 249. Identified as Menapian, Carausius cannot easily be named an
alumnus of the Franks.
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 17

Given the fact that a few decades earlier Postumus had employed troops of Germanic origin and
may have offered them settlement, both passages can be interpreted as examples of forced
deportation by the imperial authorities of groups living in Batavia.
65
The Batavian and
Cananefatian communities may have made common cause with the Frankish newcomers and
consequently received the same punishment from the Roman authorities, namely deportation as
laeti to interior Gaul. One may even go a step further and speculate that the praefecti laetorum
Batavorum stationed near Arras, Neumagen and Lyon mentioned in the Notitia Dignitatum
refer to groups of Batavians and Franks who had been deported as laeti to uninhabited regions
of interior Gaul.
66
Future research will hopefully clarify whether the fourth-century
Germanic-type settlements excavated in northern France
67
can be linked to the historically
documented deportations of Germanic groups from Germania inferior.
68
We conclude that the large number of excavations and the improved dating methods that have
become available in recent decades are now allowing us to detect recurrent patterns in site
developments. These enable us to add arguments to existing historical debates and to suggest
new hypotheses. Even though not all of the above can be proven to the letter, it is clear that
one or several catastrophic events in the late third century, in combination with imperial policy
and probably environmental problems, led to the dramatic depopulation of the countryside in
large parts of Germania inferior.
GERMANIC SETTLEMENT AND DECLINING IMPERIAL POWER
The historical developments in the fourth and early fifth century in Germania secunda were
determined by two key themes. The first is the continuous attempt of the Roman authorities to
control the Rhine corridor in order to protect the Gallic hinterland against Germanic raiders and
to keep open the strategic routes to Britannia via the Rhine and Meuse. The military history of
the late third to fifth century can best be characterised as a period during which times of
strength, when Roman military influence was restored in Lower Germany, alternated with times
of weakness or even collapse of the limes.
69
As a rule, diminished military attention given to
the northern regions led directly to incursions and pillaging by Frankish war bands, while
periods of recovery were linked to successful campaigns against Frankish groups by Postumus
(c. 260), Aurelian (c. 275), Probus (c. 276–80), Constantius (c. 293–97), Constantine I (c. 310)
and Julian II (c. 359).
70
The notion of the limes as a defensive infrastructure along the Rhine
remained alive throughout the fourth century, although little of this was realised in practice.
Material dating to the early and mid-fourth century is scarce in, if not virtually absent from, the
forts. Most castella were garrisoned incidentally at best.
71
This period is characterised by
Roman field campaigns in times of threat rather than continually manned fortifications.
Absolute low points for the limes were Stilicho’s withdrawal of regular Roman troops from the
65
Heeren 2015, 290–4. The label ‘Batavia’in the written sources may extend outside the area that we know to be
the Batavian civitas, to include parts of the sandy soils to the south of it, as well as the Cananefatian civitas (De Bruin
2019, 238).
66
Not. Dign. occ. 42. See also the discussion in Willems 1984, 275.
67
Kasprzyk 2018a, 261–6.
68
Lenz 2005.
69
For an overview of the military history of the Lower Rhine region in the late Roman period, see the excellent
thesis of De Boone 1954; cf. also Zöllner 1970,1–43; Willems 1984, 248–50 (largely relying on De Boone).
70
De Boone 1954; Zöllner 1970.
71
In past decades, incidental finds from military camps were too readily used to suggest continuity of occupation. A
dendrochronologically dated fourth-century phase of the Valkenburg castellum was often cited. However, it was
discovered recently that the dendrochronological research was faulty and did not support the dating (Rien Polak,
pers. comm.).
NICO ROYMANS ET AL.18

Rhine to Italy in 401/02 and the Germanic incursions of 405 or 406 near Mainz. The appearance of
the usurper Constantine III (407–11), who is credited by Zosimus as being the last Roman general
to restore the Rhine frontier,
72
represents a short upheaval, after which the limes quickly ceased to
be a defensive infrastructure of any real significance.
The second theme, and the most important for this paper, concerns the substantial influx of
Germanic groups, or more specifically the groups described as Franks in the written sources,
from the period around 400 onwards. Franci is a collective name for a series of smaller tribes
in the areas east and north of the Lower Rhine who had long maintained relations with the
Roman Empire. Not until the middle of the third century does this name appear in the written
accounts.
73
In the third and fourth centuries Franks were generally described as people living
outside the Roman Empire, but in the late fourth and fifth centuries they also inhabited land in
Germania secunda. Frankish society underwent a major transformation during the late Roman
period, which was closely tied to increasing interaction –both friendly and hostile –with the
Roman Empire. Viewed from this perspective, the Franks can be regarded as a product of the
late Roman frontier.
Archaeological research over the course of the last two decades has contributed considerably to
the discussion about Germanic immigrant groups, most notably the Franks. Historical accounts
about Franks settling in the Roman province date to the later third and fourth centuries, but, so
far, hardly any tangible remains have been recovered. This also applies to the report that Salian
Franks were beaten by Julian II in 359 after they settled in Toxandria.
74
It is generally
understood that they were permitted to settle, but so far there is no archaeological evidence
dating to this period. By contrast, there is now ample evidence for immigration dating to the
late fourth/early fifth century, visible in house plans, sunken-featured secondary buildings and
numerous mobile finds (FIG.9and TABLE 1). Remarkably, these settlements have been found in
areas that were severely depopulated earlier, mainly in the Meuse valley and the sandy area
south and west of the Meuse river stretching as far as the Scheldt valley. In most instances,
older settlement locations were used again, after a habitation discontinuity of more than a century.
Migration as an explanation for regional change in the late Roman period merits some elaboration
since it has been critically evaluated previously. Archaeologists based narratives of migration on
written sources and the etnische Deutung of material culture from burial contexts.
75
Critical
evaluations came from more theoretically oriented scholars, such as Guy Halsall and Frans
Theuws, who argued against any ethnic meaning of material culture and explained differences in
burial rites in other ways.
76
Although their criticism is broadly correct, their general conclusion –
that no Germanic immigration is proven –is not followed by the present authors. Apart from the
metal finds, which are accompanied by handmade pottery of Rhine-Weser-Germanic style, the
settlement complexes show house plans of a type that was common north of the Rhine and
associated with sunken-featured buildings; furthermore, rye is found, a cereal also originating
north of the Rhine as a cultivated crop.
77
Because these various classes of evidence all point to
the Germanic origin of the new settlers, we argue that archaeology can prove migration.
78
Additional archaeological information comes from gold finds, which provide more historical
background to this renewed immigration in the (former) Roman province (FIG.10).
79
There is a
72
Zos. 6.3.3 reports that Constantine ‘let rule a total security along the Rhine, which has been neglected since the
age of Julian’(taken from the French edition of Paschoud 1989,7–8, and translated into English by the present authors).
73
De Boone 1954; Zöllner 1970; Taayke 2003.
74
Amm. Marc. 17.8.
75
For instance, Böhme 1974;1999; Quast 2009.
76
Halsall 2000;2007; Theuws 2009.
77
Heeren 2017; Heeren and Roymans 2018.
78
For the broader debate on archaeology and migration, see Burmeister 2000;2017.
79
Roymans 2017.
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 19

true gold-hoard horizon in the region and many of the coin-dated hoards have a terminal coin of
Honorius (A.D. 395–423) or Constantine III (A.D. 407–11).
80
Constantine III concluded treaties
with Frankish groups and the same is hinted at by the sources about Stilicho, magister militum
of Honorius, when he withdrew the regular Roman troops to the south (401–02). In both
instances we can surmise that these Franks were allowed to settle in the former province, in
return for military service. This settlement took place in uninhabited areas where the troops
received the status of foederati, not laeti. The free, federate status of the new settlers is
FIG. 9. Distribution of excavated Germanic settlements from the late fourth and early fifth centuries in Germania
secunda. The numbering corresponds with the numbering of sites in TABLE 1.
80
Roymans 2017, fig. 3.
NICO ROYMANS ET AL.20

TABLE 1. LIST OF EXCAVATED GERMANIC SETTLEMENTS FROM THE LATE FOURTH/EARLY FIFTH
CENTURY IN GERMANIA SECUNDA
(The numbering corresponds with the numbering of sites in FIG.9(after Heeren 2017). SFB = sunken featured building.
References for the settlements are available in Heeren 2015;2017; Van Enckevort et al.2017)
No Site Settlement features and date Rye
1 Nijmegen-Sint Canisiussingel 1 SFB; charred; late Roman
2 Wijchen-Tienakker Former villa, burgus site 4th century, settlement (8 SFBs); late
4th/5th century
3 Tiel-Passewaaij Two shorthouses, 1 of Wijster B2 type; late 4th/5th century
4 Valburg-Molenzicht 2 longhouses (1 possibly Wijster type), 6 wells; late 4th/5th century
5 Naaldwijk-Hoogwerf Former vicus, 1 SFB, 1 well; late 4th/5th century
6 Alphen-Kerkakkers 28 SFBs, 9 main houses (no Wijster type); late 4th to early 6th century Yes
7 Asten-Prins Bernhardstraat 1 SFB, coarse pottery; late Roman and/or medieval
8 Breda West-Steenakker 6 or 8 SFBs, 2 longhouses of which 1 Wijster type; around A.D. 400
9 Cuijk-Heeswijkse Kampen At least 1 SFB; 3rd century or later
10 Cuijk-De Nielt 2 main buildings (no Wijster type) early 4th century; 6 SFBs late
4th century
11 Geldrop-’t Zand 13 SFBs, 2 main buildings; late 4th to mid-5th century Yes
12 Goirle-Huzarenwei 2 SFBs, 4 main buildings (3 Wijster type); late 4th/early 5th century Yes
13 Lierop-Steemertseweg 1 SFB; undated; near late Roman pits
14 Tilburg-Stappegoor 2 shorthouses; first half of 5th century
15 Buggenum-Wijnaerden Several Wijster buildings and SFBs; late 4th/5th century
16 Gennep-Stamelberg More than 100 SFBs, at least 12 Wijster-like longhouses; late
4th/5th century
Yes
17 Grubbenvorst-De Soom 9 SFBs, 2 main buildings; some late 4th/early 5th century, some
medieval
18 Helden-Schrames 4 SFBs, 6 shorthouses Wijster B2, 1 Wijster longhouse; late 4th/5th century
19 Horst-Hoogveld 2 SFBs; apparently 3rd century
20 Maastricht-Witmakersstraat 1 SFB; mid-5th century
21 Holtum-Noord 11 longhouses (no Wijster type), 8 SFBs; late 4th/5th century
22 Swalmen 1 late Roman SFB next to Merovingian cemetery
23 Blerick-Heierhoeve Around 10 SFBs; late 1st/2nd century settlement, also 4th-century coins
24 Voerendaal-Ten Hove At least 17 SFBs at villa site; 4th to 7th century Yes
25 Baelen-Nereth 1 longhouse, 1 shorthouse, 4 possible SFBs; mid- to late 4th century
26 Neerharen-Rekem At least 25 SFBs, 2 Wijster-like longhouses; late 4
th
to early 5th century
27 Wange-Damekot 7 SFBs at late Roman villa site; 5th to 6th century
28 Herk-de-Stad, Donk 1 SFB or byre section of longhouse(?); mid- to late 4th century
29 Meldert-Zelemsebaan 2 SFBs and unclear main buildings; late 4th/5th century Yes
30 Sint-Martens-Latem 1 SFB; late 4th/early 5th century
31 Nazareth-’s-Gravendreef Longhouse Wijster A type; late 3rd century(?)
32 Froitzheim 1 SFB within former burgus; c. A.D. 300(?)
33 Harff 2 SFBs on former villa terrain; second half of 4th century
34 Rodenkirchen 1 SFB on former villa terrain; 4th century Yes
35 Ede-Veldhuizen 180 main houses and large buildings, 4 SFBs, 208 small outbuildings,
67 wells
36 Ede-Op den Berg 80+ main houses and/or large outbuildings, 65 small outbuildings,
many SFBs
37 Ede-Uitvindersbuurt 2 longhouses, 2 shorthouses, 2 possible SFBs, 3 to 6 wells
38 Ede-Bennekom Several longhouses and at least 4 SFBs
39 Leersum-Middenweg Buildings, SFBs; late Roman/early Middle Ages
40 Didam-Aalsbergen/Kollenburg 14 main buildings, more than 30 SFBs, 4 small outbuildings, 26 wells
41 Didam-Kerkwijk 2 three-aisled main buildings, 4 wells
42 Wehl-Hessenveld 7 longhouses, some of Wijster type, 11 partial other houses, over 20 SFBs
43 Wehl-Oldershove 1 Wijster B2 shorthouse, 2 SFBs; close to Hessenveld
44 Zutphen-Ooijershoek c. 20 SFBs and several Wijster A longhouses
45 Deventer-Colmschate 10 or 12 SFBs Yes
Continued
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 21

TABLE 1. CONTINUED
No Site Settlement features and date Rye
46 Bathmen-Bergakker Some SFBs; one with a confirmed 4th-century date
47 Heeten-Hordelman 39 SFBs Yes
48 Heeten-Telgen 14 SFBs, longhouses
49 Markelo-Elsen 1 SFB, pits, a well (trial trench)
50 Wierden-Enter 3 SFBs, longhouse
51 Denekamp-De Borchert 11 SFBs, various longhouses
52 Emmen-Noordbarge c. 35 SFBs, some longhouses
53 Emmen-Frieslandweg 8 SFBs, longhouses
54 Coevorden-Diphoorn 1 SFB, longhouses
55 Wijster c. 85 SFBs, many Wijster longhouses and shorthouses
56 Midlaren-De Bloemert 27 SFBs, various types of longhouses
57 Leeuwarden-Oldehoofsterkerkhof Several Wijster-type longhouses
58 Wijnaldum-Tjitsma Several Wijster-type longhouses
59 Castricum-Oosterbuurt Several Wijster-type longhouses
60 Castrop-Rauxel, Ickern 2 Wijster-type longhouses
FIG. 10. Distribution of late Roman solidi (A.D. 364–455) in the Lower Rhine frontier zone (after Roymans 2017,
fig. 1, with additions).
NICO ROYMANS ET AL.22

supported by the study of the extensive gold circulation and deposition in this period (FIG.10). The
large-scale payments of Roman gold in the form of solidi and ornaments to Frankish groups
illustrate the shifting balance of power on the late Roman frontier.
We conclude that there was a partial repopulation of the countryside around the turn of the
fourth to the fifth century. This model of substantial Germanic immigration to the Roman
province is particularly valid for former substantially depopulated areas, but this may also be
due to the greater archaeological visibility of immigration in those areas. The motor behind the
gold payments and land allotments to Germanic groups was the control over the war bands
recruited from these rural groups. What is new in this migration discussion is the combined
approach analysing historical sources, house forms, food remains, jewellery and pottery. The
next step currently underway is to combine these data with the isotopic analysis of animal teeth
(because livestock may have been brought on the hoof in the case of first-generation settlers)
and the analysis of burial customs in relation to the isotopic signature of human remains (both
inhumed and cremated).
CONCLUSIONS AND DISCUSSION
The case study of Germania inferior presented above teaches us that rural populations were
generally integrated more closely into Roman imperial structures than is often assumed.
Geographically, they may have been far removed from the urban or military centres, but
imperial power networks extended into the furthest corners of every civitas. In Lower Germany
rural groups had to contend not only with taxation and recruitment systems, but, in situations
of crisis, also with extreme mass violence, land expropriation and forced deportations. Central
to this perspective is the understanding of power relations, including their economic and
religious dimensions. Archaeology has the potential to explore further and operationalise this
theme in regional research programmes.
Certain issues keep recurring in the analysis outlined above. The first is human mobility. In the
first two centuries A.D. this was primarily individual mobility, linked to service in the Roman army.
In other periods it mainly involved group migrations associated with demographic rearrangements
of rural populations in a region. A second recurring theme is the intense connectivity between rural
communities and the Roman military domain. Certainly in a heavily militarised frontier province,
rural developments cannot be understood without considering the multifaceted networks with the
military community. A third recurring theme is the phenomenon of the appearing and disappearing
ethnic groups of the Roman frontier zone. The written sources suggest that most frontier tribes
were to some extent new political creations; this is supported by archaeological research
showing that tribal units often lacked a homogenous material footprint in the spheres of pottery
style, house architecture and ornaments.
By far the most dominant narrative in Roman rural archaeology is that of the progressive
romanisation of groups, thereby focusing on the rise of villa landscapes, a monetary economy,
a market-oriented production and new styles of consumption. From the perspective employed in
this study we present an alternative series of topics embedded in a different narrative, which are
only marginally discussed in most archaeological syntheses, but which should not be absent
from a balanced approach to developments in the Roman countryside. One might even go a
step further and state that the developments discussed in this paper must have been experienced
as dramatic episodes by the rural communities themselves.
The above insights obtained for the Lower Germanic region also lend themselves to comparison
with developments in other provinces. The recent synthesis available for Roman Britain shows
important differences on some main points. Alexander Smith and Michael Fulford conclude
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 23

that the Roman conquest had no significant impact on existing settlement patterns in Britannia; the
general pattern was one of continuity of rural settlement from the late Iron Age to the later second
century A.D.
81
Also absent from Britannia are indications of a substantial influx of external groups
in the early post-conquest period, and ethnic recruitment there certainly did not reach the high
levels attested for Germania inferior. Finally, the process of settlement abandonment and
depopulation of the countryside in the late Roman period followed a more gradual pattern
without a dramatic low point in the late third century.
82
The recent synthesis presented by
Reddé and colleagues for interior Gaul and more specifically the province of Gallia Belgica
also shows differences from the patterning sketched for Germania inferior. In many regions
there is evidence for a substantial population decline in the La Tène D period, but here the
Roman conquest seems to have played a secondary role at most, since the trend had already
started in the late second century B.C.
83
A gradual increase in the population is observed from
the Augustan period onward, but the immigration of tribal groups did not play any role there.
84
The level of ethnic recruitment in interior Gaul is not discussed, but according to the historical
sources it must have been of limited importance (see FIG.6). As in Britannia, the rural
settlement pattern seems to have remained largely intact well into the fourth century.
85
There is
some evidence for the appearance of Germanic-type settlements in the Seine valley in the
fourth century, but it is too early to draw any conclusions about the relative extent of this
phenomenon.
This comparison along main lines with the evidence from Britannia and Gallia Belgica teaches
us that the developments sketched for Germania inferior are largely specific to that area and cannot
simply be extrapolated to other provinces. Each province has its own story to tell, which for
Germania inferior is determined above all by the continuous and multifaceted interaction with
the immense territory and many peoples of Germania magna. In all provinces, however, rural
communities were deeply embedded in the power structures of the Roman Empire.
Power-related themes affecting rural communities –such as warfare, mass violence, deportation
and economic exploitation and repression –are increasingly marginalised in current research,
which often focuses on the ‘soft’, cultural dimensions of the transformation of the rural world.
Finally, the analysis above shows us that written sources can generate interesting hypotheses for
archaeological research and that, vice versa, archaeological insights may lead to different
interpretations of written sources or draw our attention to major biases in the historical record.
There is an abundance of themes for combined historical-archaeological research, and the
constant increase in the quantity and quality of archaeological data allows us to ask more
complex questions. In contrast to the situation a few generations ago, archaeology no longer
participates in this debate as an historical sub-discipline, but as an autonomous discipline which
has its own methodologies and generates its own empirical data. We can look forward to a
phase of renewed interdisciplinary cooperation.
ACKNOWLEDGEMENTS
The authors wish to thank Andrew Lawrence (University of Bern) and the anonymous reviewer of Britannia
for their comments on an earlier draft of this paper, and Bert Brouwenstijn for the cartography and
illustrations. The English was checked by Annette Visser (New Zealand). This study is part of the
research output of the projects Portable Antiquities of the Netherlands and Tiel-Medel as a Key Site for
81
Smith and Fulford, in Smith et al. 2016, 408–10.
82
Smith and Fulford, in Smith et al. 2016, 414.
83
Malrain et al.2013; Reddé 2018b, 135.
84
Reddé 2018a, 498.
85
Kasprzyk 2018b.
NICO ROYMANS ET AL.24

Innovative Research Towards Migration and Ethnogenesis in the Roman Frontier. Both projects are funded by
the Netherlands Organisation for Scientific Research and are run by the Archaeological Department of Vrije
Universiteit Amsterdam.
Department of Archaeology, Classics and Ancient Studies, Vrije Universiteit Amsterdam
(N.R., T.D. and S.H.)
n.g.a.m.roymans@vu.nl
a.m.j.derks@vu.nl
s.heeren@vu.nl
BIBLIOGRAPHY
Alföldy, G. 1968: Die Hilfstruppen in der römischen Provinz Germania inferior, Epigraphische Studien 6,
Düsseldorf
Allen, M., Lodwick, L., Brindle, T., Fulford, M., and Smith, A. (eds) 2017: The Rural Economy of Roman
Britain, Britannia Monograph 30, London
Andrews, C. 2013: ‘Are Roman seal-boxes evidence for literacy?’,Journal of Roman Archaeology 26, 423–38
Andrikopoulou-Strack, J.-N., Enzenberger, P., Frank, K., Keller, C., and Klän, N. (eds) 1999: ‘Eine
frührömische Siedlung in Jüchen-Neuholz: Überlegungen zur Siedlungskontinuität in der Lößbörde’,
Bonner Jahrbücher 199, 141–80
Bechert, T. 1982: Römisches Germanien, Munich
Bloemers, J.H.F. 1978: Rijswijk (Z.H.), ‘De Bult’: eine Siedlung der Cananefaten I–III, Amersfoort
Böhme, H.W. 1974: Germanische Grabfunde des 4. und 5. Jahrhunderts zwischen unterer Elbe und Loire:
Studien zur Chronologie und Bevölkerungsgeschichte, Münchner Beiträge zur Vor- und Frühgeschichte 19,
Munich
Böhme, H.W. 1999: ‘Sächsische Söldner im römischen Heer: das Land zwischen Ems und Niederelbe
während das 4. und 5. Jahrhundert’, in F. Both and H. Aouni (eds), Über allen Fronten:
Nordwestdeutschland zwischen Augustus und Karl dem Grossen, Archäologische Mitteilungen aus
Nordwestdeutschland Beiheft 26, Oldenburg, 49–73
Bowman, A.K., and Thomas, J.D. 1994–2003: The Vindolanda Writing Tablets: Tabulae Vindolandenses
2–3, London
Breeze, D. 2018: ‘The value of studying Roman frontiers’,Theoretical Roman Archaeology Journal 1.1,
https://doi.org/10.16995/traj.212
Bridger, C. 2003: ‘Das spätantike Xanten: eine Bestandsaufnahme’, in T. Grünewald and S. Seibel (eds),
Kontinuität und Diskontinuität: Germania inferior am Beginn und am Ende der römischen Herrschaft,
Berlin and New York, 12–36
Brüggler, M., Jeneson, K., Gerlach, R., Meurers-Balke, J., Zerl, T., and Herchenbach, M. (eds) 2017: ‘The
Roman Rhineland: farming and consumption in different landscapes’, in M. Reddé (ed.), Gallia Rustica
1: les campagnes du nord-est de la Gaule, de la fin de l’âge du Fer à l’Antiquité tardive, Ausonius
Éditions Mémoires 49, Bordeaux, 19–96
Burmeister, S. 2000: ‘Archaeology and migration: approaches to an archaeological proof of migration’,
Current Anthropology 41.4, 539–67
Burmeister, S. 2017: ‘One step beyond: Migration als kulturelle Praxis’, in A. Dietz, A. Hidding and
J.D. Preisigke (eds), Migration and Change: Causes and Consequences of Mobility in the Ancient
World, Distant Worlds Journal 3, Munich, 3–18
Clotuche, R., Derreumaux, M., Pigière, F., Jouanin, G., Preiss, S., and Yvinec, J.-V. (eds) 2017: ‘Les
campagnes du territoire nervien: approches croisées’, in M. Reddé (ed.), Gallia Rustica 1: les
campagnes du nord-est de la Gaule, de la fin de l’âge du Fer à l’Antiquité tardive, Ausonius Éditions
Mémoires 49, Bordeaux, 179–210
De Blois, L. 2017: ‘Invasions, deportations, and repopulation: mobility and migration in Thrace, Moesia
Inferior, and Dacia in the third quarter of the third century AD’, in E. Lo Cascio and L.E. Tacoma
(eds), The Impact of Mobility and Migration in the Roman Empire, Impact of Empire 22, Leiden and
Boston, 42–54
De Boone, W.J. 1954: De Franken van hun eerste optreden tot de dood van Childerik, Groningen
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 25

De Bruin, J. 2019: Border Communities at the Edge of the Roman Empire: Processes of Change in the Civitas
Cananefatium, Amsterdam Archaeological Studies 28, Amsterdam
De Jonge, W., Bazelmans, J., and de Jager, D. (eds) 2006: Forum Hadriani: van Romeinse stad tot
monument, Utrecht
Degler, A. 2017: ‘Aureus of Postumus with the owner’s graffito (Gounthiou)’,Proceedings XV: International
Numismatic Congress Taormina 2015, Taormina, 723–6
Derks, T. 2009: ‘Ethnic identity in the Roman frontier: the epigraphy of Batavi and other Lower Rhine tribes’,
in T. Derks and N. Roymans (eds), Ethnic Constructs in Antiquity: The Role of Power and Tradition,
Amsterdam Archaeological Studies 13, Amsterdam, 239–82
Derks, T., and Roymans, N. 2002: ‘Seal-boxes and the spread of Latin literacy in the Rhine delta’,in
A.E. Cooley (ed.), Becoming Roman, Writing Latin? Literacy and Epigraphy in the Roman West,
Journal of Roman Archaeology Supplement 48, Porthsmouth, RI, 87–134
Drinkwater, J. 2008: ‘Maximinus to Diocletian and the “crisis”’, in A.P. Bowman, P. Garnsey and
A. Cameron (eds), The Cambridge Ancient History 12: The Crisis of Empire AD 193–337, Cambridge,
28–66
Eck, W. 2004: Köln in römischer Zeit: Geschichte einer Stadt im Rahmen des Imperium Romanum, Cologne
Eck, W. 2008: ‘Die Gründung der Colonia Ulpia Traiana in ihrem politische Kontext’, in M. Müller,
H.J. Schalles and N. Zieling (eds), Colonia Ulpia Traiana: Xanten und sein Umland in römischer Zeit,
Mainz, 243–55
Eckardt, H., and Müldner, G. 2016: ‘Mobility, migration, and diaspora in Roman Britain’, in M. Millett,
L. Revell and A. Moore (eds), The Oxford Handbook of Roman Britain, Oxford, 202–23
Erdrich, M. 1998: ‘Terra Nigra-Fussschalen wie Chenet 342 oder Gellep 273: eine salisch-fränkische
Keramikgattung’,Germania 76–2, 875–84
Frank, K. 2018: ‘Mehrum: neues zu einem alten Grabfund’, in M. Aufleger and P. Tutlies (eds), Das Ganze
ist mehr als die Summe seiner Teile: Festschrift für Jürgen Kunow, Bonn, 465–76
Gaitzsch, W. 2011: ‘Roman villa landscapes of the lignite mining areas in the hinterland of Cologne’,in
N. Roymans and T. Derks (eds), Villa Landscapes in the Roman North: Economy, Culture and
Lifestyles, Amsterdam Archaeological Studies 17, Amsterdam, 285–300
Galsterer, H. 1990: ‘Von den Eburonen zu den Agrippinensern’,Kölner Jahrbuch 23, 117–26
Gardner, A. 2013: ‘Thinking about Roman imperialism: postcolonialism, globalisation and beyond?’,
Britannia 44, 1–25
Gechter, M., and Kunow, J. 1986: ‘Zur ländlichen Besiedlung des Rheinlandes in römischer Zeit’,Bonner
Jahrbücher 186, 377–96
Gouw-Bouman, M.T.I.J., van Asch, N., Engels, S., and Hoek, W.Z. 2019: ‘Late Holocene ecological shifts
and chironomid-inferred summer temperature changes reconstructed from Lake Uddelermeer, the
Netherlands’,Palaeogeography, Palaeoclimatology, Palaeoecology 535, 109366, https://doi.org/10.
1016/j.palaeo.2019.109366
Groenman-Van Waateringe, W. 1983: ‘The disastrous effect of the Roman occupation’, in R. Brandt and
J. Slofstra (eds), Roman and Native in the Low Countries: Spheres of Interaction, BAR International
Series 184, Oxford, 147–57
Haas, J. 2006: Die Umweltkrise des 3. Jahrhunderts n. Chr. im Nordwesten des Imperium Romanum:
interdisziplinäre Studien zu einem Aspekt der allgemeinen Reichskrise im Bereich der beiden
Germaniae sowie der Belgica und der Raetia, Geographica Historica 22, Stuttgart
Halsall, G. 2000: ‘Archaeology and the late Roman frontier in northern Gaul: the so-called “Föderatengräber”
reconsidered’, in W. Pohl and H. Reimitz (eds), Grenze und Differenz im frühen Mittelalter, Forschungen
zur Geschichte des Mittelalters 1/Österreichische Akademie der Wissenschaften, Philosophisch-Historische
Klasse Denkschriften 287, Vienna, 167–80
Halsall, G. 2007: Barbarian Migrations and the Roman West 376–568 (2nd edn), Cambridge
Haselgrove, C. 1984: ‘Warfare and its aftermath as reflected in the precious metal coinage of Belgic Gaul’,
Oxford Journal of Archaeology 3.1, 81–105
Haselgrove, C. 2019: ‘The Gallic War in the chronology of Iron Age coinage’, in A. Fitzpatrick and
C. Haselgrove (eds), Julius Caesar’s Battle for Gaul: New Archaeological Perspectives, Oxford and
Philadelphia, 241–66
NICO ROYMANS ET AL.26

Heeren, S. 2009: Romanisering van rurale gemeenschappen in de civitas Batavorum: de casus
Tiel-Passewaaij, Nederlandse Archeologische Rapporten 36, Amersfoort
Heeren, S. 2015: ‘The depopulation of the Lower Rhine region in the third century: an archaeological
perspective’, in N. Roymans, T. Derks and H.A. Hiddink (eds), The Roman Villa of Hoogeloon and the
Archaeology of the Periphery, Amsterdam Archaeological Studies 22, Amsterdam, 271–94
Heeren, S. 2017: ‘From Germania Inferior to Germania secunda and beyond: a case study of migration,
transformation and decline’, in N. Roymans, S. Heeren and W. De Clercq (eds), Social Dynamics in the
Northwest Frontiers of the Late Roman Empire: Beyond Decline or Transformation, Amsterdam
Archaeological Studies 26, Amsterdam, 149–78
Heeren, S., and Roymans, N. 2018: ‘Contextualising ethnicity and the rhetoric of burial rites in late antique
northern Gaul: the evidence from settlements and precious metal circulation’, in M. Kars, R. van Oosten,
M.A. Roxburgh and A. Verhoeven (eds), Rural Riches and Royal Rags? Studies on Medieval and Modern
Archaeology, Presented to Frans Theuws, Zwolle, 8–14
Heimberg, U. 2002–03: ‘Römische Villen an Rhein und Maas’,Bonner Jahrbücher 202–3, 57–148
Heinrichs, J. 2001: ‘Römische Perfidie und germanischer Edelmut? Zur Umsiedlung protocugernischer
Gruppen in den Raum Xanten 8 v.Chr.’, in T. Grünewald (ed.), Germania inferior: Besiedlung,
Gesellschaft und Wirtschaft an der Grenze der römisch-germanischen Welt, Berlin and New York, 54–92
Hingley, R. 2005: Globalizing Roman Culture: Unity, Diversity and Empire, London
Hornung, S. 2016: Siedlung und Bevölkerung in Ostgallien zwischen Gallischem Krieg und der Festigung der
römischen Herrschaft, Römisch-Germanische Forschungen 73, Darmstadt and Mainz
Huskinson, J. (ed.) 2000: Experiencing Rome: Culture, Identity and Power in the Roman Empire, London
Kalis, A.J., Karg, S., Meurers-Balke, J., and Teunissen-van Oorschot, H. 2008: ‘Mensch und Vegetation am
unteren Niederrhein während der Eisen- und Römerzeit’, in M. Müller, H.J. Schalles and N. Zieling (eds),
Colonia Ulpia Traiana: Xanten und sein Umland in römischer Zeit, Mainz, 31–48
Kalis, A.J., and Meurers-Balke, J. 2007: ‘Landnutzung im Niederrheingebiet zwischen Krieg und Frieden’,in
M. Hegewisch (ed.), Krieg und Friede: Kelten Römer Germanen, Bonn, 144–53
Kasprzyk, M. 2018a: ‘L’équipement et les formes des établissements ruraux du Bas-Empire (milieu du IIIe –
milieu du Ve siècle p.C.) dans l’est du diocèse des Gaules’, in M. Reddé (ed.), Gallia Rustica 2: les
campagnes du nord-est de la Gaule, de la fin de l’âge du Fer à l’Antiquité tardive, Bordeaux, 235–305
Kasprzyk, M. 2018b: ‘Les mutations de l’Antiquité tardive dans l’est du diocèse des Gaules’, in M. Reddé
(ed.), Gallia Rustica 2: les campagnes du nord-est de la Gaule, de la fin de l’âge du Fer à l’Antiquité
tardive, Bordeaux, 611–35
Krebs, C.B. 2006: ‘“Imaginary geography”in Caesar’sBellum Gallicum’,American Journal of Philology
127.1, 111–36
Krebs, C.B. 2011: ‘Borealism: Caesar, Seneca, Tacitus and the Roman discourse about the Germanic north’,in
E. Gruen (ed.), Cultural Identity in the Ancient Mediterranean: Issues and Debates, Los Angeles, 202–21
Krebs, C.B. 2018: ‘The world’s measure: Caesar’s geographies of Gallia and Britannia in their contexts and
as evidence of his world map’,American Journal of Philology 139, 93–122
Lenz, K.H. 1999: Siedlungen der römischen Kaiserzeit auf den Aldenhovener Platte, Rheinische
Ausgrabungen 45, Cologne and Bonn
Lenz, K.H. 2005: ‘Germanische Siedlungen des 3. bis 5. Jahrhunderts n.Chr. in Gallien: schriftliche
Überlieferung and archäologische Befunde’,Berichte der Römisch-Germanischen Kommission 86, 349–
444
Lenz, K.H. 2006: ‘Veteranen der römischen Armee im Siedlungsbild einer früh- und mittelkaiserzeitlichen
Koloniestadt und deren Hinterland: das Beispiel der Colonia Claudia Ara Agrippinensium (Köln)’,
Germania 84, 61–91
Malrain, F., Blancquart, G., and Lorho, T. 2013: L’habitat rural du Second Âge du Fer: rythmes de création
et d’abandon au nord de la Loire, Paris
Mattingly, D. 2006: An Imperial Possession: Britain in the Roman Empire 54 BC–AD 409, London
Mattingly, D.J. 2011: Imperialism, Power and Identity: Experiencing the Roman Empire, Princeton and
Oxford
McCormick, M., Büntgen, U., Cane, M., Cook, E.R., Harper, P., Huybers, P., Litt, T., Manning, S.W.,
Mayewski, P.A., More, A., Nicolussi, K., and Tegel, W. 2012: ‘Climate change during and after the
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 27

Roman Empire: reconstructing the past from scientific and historical evidence’,Journal of Interdisciplinary
History 43.2, 169–220
Nash, D. 1987: ‘Imperial expansion under the Roman Republic’, in M. Rowlands, M. Trolle Larsen and
K. Kristiansen (eds), Centre and Periphery in the Ancient World, Cambridge, 87–103
Nicolay, J. 2007: Armed Batavians: Use and Significance of Weaponry and Horse Gear from Non-Military
Contexts in the Rhine Delta (50 BC–AD 450), Amsterdam Archaeological Studies 11, Amsterdam
Nixon, C.E.V., and Rodgers, B.S. 1994: In Praise of Later Roman Emperors: The Panegyrici Latini:
Introduction, Translation and Historical Commentary, Transformation of the Classical Heritage 21,
Berkeley
Otten, T., and Ristow, S. 2008: ‘Xanten in der Spätantike’, in M. Müller, H.J. Schalles and N. Zieling (eds),
Colonia Ulpia Traiana: Xanten und sein Umland in römischer Zeit, Mainz, 549–82
Päffgen, B. 2012: ‘Köln und sein Umland zur Zeit der Soldatenkaiser (235–285 n.Chr.), besonders im
Hinblick auf das Gallische Sonderreich’, in T. Fischer (ed.), Die Krise des 3. Jahrhunderts n.Chr. und
das Gallische Sonderreich: Akten des interdisziplinären Kolloquiums Xanten 2009, Wiesbaden, 97–150
Paschoud, F. 1989: Zosime: Histoire nouvelle III-2 (livre VI et index), Paris
Quast, D. 2009: Foreigners in Early Medieval Europe: Thirteen International Studies on Early Medieval
Mobility, Monographien des Römisch-Germanischen Zentralmuseums 78, Mainz
Raepsaet, G. 2013: ‘L’ethnogenèse de la civitas Tungrorum et la formation de la province de Germanie’,
L’Antiquité classique 82, 11–148
Raepsaet-Charlier, M.-T. 2002–03: ‘Vielfalt und kultureller Reichtum in den civitates Niedergermaniens’,
Bonner Jahrbücher 172, 35–56
Reddé, M. (ed.) 2017: Gallia Rustica 1: les campagnes du nord-est de la Gaule, de la fin de l’âge du Fer à
l’Antiquité tardive, Bordeaux
Reddé, M. (ed.) 2018a: Gallia Rustica 2: les campagnes du nord-est de la Gaule, de la fin de l’âge du Fer à
l’Antiquité tardive, Bordeaux
Reddé, M. 2018b: ‘The impact of the German frontier on the economic development of the countryside of
Roman Gaul’,Journal of Roman Archaeology 31, 131–60
Riechelmann, D.F.C., and Gouw-Bouman, M.T.I.J. 2019: ‘A review of climate reconstructions from
terrestrial climate archives covering the first millennium AD in northwestern Europe’,Quaternary
Research 91.1, 111–31
Riggsby, A. 2006: Caesar in Gaul and Rome: War in Words, Austin
Roymans, N. 2004: Ethnic Identity and Imperial Power: The Batavians in the Early Roman Empire,
Amsterdam Archaeological Studies 10, Amsterdam
Roymans, N. 2009: ‘Hercules and the construction of a Batavian identity in the context of the Roman empire’,
in T. Derks and N. Roymans (eds), Ethnic Constructs in Antiquity: The Role of Power and Tradition,
Amsterdam Archaeological Studies 13, Amsterdam, 219–38
Roymans, N. 2011: ‘Ethnic recruitment, returning veterans and the diffusion of Roman culture among rural
populations in the Rhineland frontier zone’, in N. Roymans and T. Derks (eds), Villa Landscapes in the
Roman North: Economy, Culture and Lifestyles, Amsterdam Archaeological Studies 17, Amsterdam,
139–60
Roymans, N. 2017: ‘Gold, Germanic foederati and the end of imperial power in the late Roman north’,in
N. Roymans, S. Heeren and W. De Clercq (eds), Social Dynamics in the Northwest Frontiers of the
Late Roman Empire: Beyond Decline or Transformation, Amsterdam Archaeological Studies 26,
Amsterdam, 57–80
Roymans, N. 2018: ‘A Roman massacre in the far north: Caesar’s annihilation of the Tencteri and Usipetes in
the Dutch river area’, in M. Fernández-Götz and N. Roymans (eds), Conflict Archaeology: Materialities of
Collective Violence in Late Prehistoric and Early Historic Europe, 167–81
Roymans, N. 2019: ‘Conquest, mass violence and ethnic stereotyping: investigating Caesar’s actions in the
Germanic frontier zone’,Journal of Roman Archaeology 32, 439–58
Roymans, N., and Derks, T. (eds) 2011: Villa Landscapes in the Roman North: Economy, Culture and
Lifestyles, Amsterdam Archaeological Studies 17, Amsterdam
Roymans, N., and Derks, T. 2015: ‘Towards a narrative of the social history of the Hoogeloon villa
settlement’, in N. Roymans, T. Derks and H.A. Hiddink (eds), The Roman Villa of Hoogeloon and the
Archaeology of the Periphery, Amsterdam Archaeological Studies 22, Amsterdam, 295–306
NICO ROYMANS ET AL.28

Roymans N., Derks, T., and Hiddink, H.A. (eds) 2015: The Roman Villa of Hoogeloon and the Archaeology
of the Periphery, Amsterdam Archaeological Studies 22, Amsterdam
Roymans, N., Heeren, S., and De Clercq, W. (eds) 2017: Social Dynamics in the Northwest Frontiers of the
Late Roman Empire: Beyond Decline or Transformation, Amsterdam Archaeological Studies 26,
Amsterdam
Roymans, N., and Scheers, S. 2012: ‘Eight gold hoards from the Low Countries: a synthesis’, in N. Roymans,
G. Creemers and S. Scheers (eds), Late Iron Age Gold Hoards from the Low Countries and the Caesarian
Conquest of Northern Gaul, Amsterdam Archaeological Studies 18, Amsterdam, 1–46
Rüger, C. 1968: Germania inferior: Untersuchungen zur Territorial- und Verwaltungsgeschichte
Niedergermaniens in der Prinzipatszeit, Beihefte Bonner Jahrbücher 30, Cologne and Graz
Schadee, H. 2008: ‘Caesar’s construction of northern Europe: inquiry, contact and corruption in De Bello
Gallico’,The Classical Quarterly 58, 158–80
Scharff, R. 2005: Der Dux Mogontiacensis und die Notitia Dignitatum, Ergänzungsbände zum Reallexikon
der Germanische Altertumskunde 50, Berlin and New York
Schulte, B. 1983: Die Goldprägung der gallischen Kaiser von Postumus bis Tetricus, Aarau
Seeck, O. 1876 (1962): Notitia Dignitatum: accedunt Notitia urbis Constantinopolitanae et laterculi
prouinciarum, Frankfurt
Slofstra, J. 2002: ‘Batavians and Romans on the Lower Rhine: the romanisation of a frontier area’,
Archaeological Dialogues 9, 16–38
Smith, A., Allen, M., Brindle, T., and Fulford, M. 2016: The Rural Settlement of Roman Britain, Britannia
Monograph 29, London
Smith, A., Allen, M., Brindle, T., Fulford, M., Lodwick, L., and Rohnbogner, A. 2018: Life and Death in the
Countryside of Roman Britain, Britannia Monograph 31, London
Taayke, E. 2003: ‘Wir nennen sie Franken und sie lebten nördlich des Rheins, 2.–5. Jh’, in E. Taayke,
J.H. Looijenga, O.H. Harsema and H.R. Reinders (eds), Essays on the Early Franks, Groningen
Archaeological Studies 1, Groningen, 1–23
Taayke, E., Peen, C., Van der Harst-Van Domburg, M., and Vos, W. 2012: Ede vol erven: Germaanse
bewoning op de rand van een wereldrijk (500 voor Chr. tot 500 na Chr.), Leiden
Theuws, F. 2009: ‘Grave goods, ethnicity, and the rhetoric of burial rites in late antique northern Gaul’,in
T. Derks and N. Roymans (eds), Ethnic Constructs in Antiquity: The Role of Power and Tradition,
Amsterdam Archaeological Studies 13, Amsterdam, 283–319
Todd, M. 2008: ‘The Germanic people and Germanic society’in A.P. Bowman, P. Garnsey and A. Cameron
(eds), The Cambridge Ancient History 12: The Crisis of Empire AD 193–337, Cambridge, 440–60
Van Driel-Murray, C. 2003: ‘Ethnic soldiers: the experience of the Lower Rhine tribes’, in T. Grünewald and
S. Seibel (eds), Kontinuität und Diskontinuität: Germania inferior am Beginn und am Ende der römischen
Herrschaft, Berlin and New York, 200–17
Van Enckevort, H., Hendriks, J., and Nicasie, M. 2017: Nieuw licht op donkere eeuwen: de overgang van de
laat-Romeinse tijd naar de vroege middeleeuwen in Zuid-Nederland, Nederlandse Archeologische
Rapporten 58, Amersfoort
Van Enckevort, H., and Thijssen, J.R.A.M. 2014: ‘Het Valkhof en omgeving tot aan het einde van de
Romeinse tijd’, in H. Peterse, D. Verhoeven, R. Camps, R. Klein, B. Kruijsen, J. Kuys, M. Nicasie and
M. Smit (eds), Het Valkhof 200 jaar geschiedenis, Nijmegen, 23–42
Van Es, W.A. 1967: Wijster: A Native Village beyond the Imperial Frontier, AD 150–425, Groningen
Vanderhoeven, A. 2017: ‘The late Roman town of Tongeren’, in N. Roymans, S. Heeren and W. De Clercq
(eds), Social Dynamics in the Northwest Frontiers of the Late Roman Empire: Beyond Decline or
Transformation, Amsterdam Archaeological Studies 26, Amsterdam, 127–48
Verhagen, J., and Heeren, S. 2016: ‘Castra Herculis: de naam van de Romeinse militaire versterking in
Nijmegen herontdekt’,Westerheem 65, 239–49
Vos, W.K. 2009: Bataafs platteland: het Romeinse nederzettingslandschap in het Nederlandse
Kromme-Rijngebied, Nederlandse Archeologische Rapporten 35, Amersfoort
Webster, J., and Cooper, N.J. (eds) 1996: Roman Imperialism, Post-Colonial Perspectives: Proceedings of a
Symposium Held at Leicester University in November 1994, Leicester archaeology Monograph 3, Leicester
Whittaker, C.R. 1994: Frontiers of the Roman Empire: A Social and Economic Study, Baltimore and London
ROMAN IMPERIALISM AND THE TRANSFORMATION OF RURAL SOCIETY 29

Whittaker, C.R. 2009: ‘Ethnic discourses on the frontiers of Roman Africa’, in T. Derks and N. Roymans
(eds), Ethnic Constructs in Antiquity: The Role of Power and Tradition, Amsterdam Archaeological
Studies 13, Amsterdam, 189–206
Willems, W.J.H. 1984: ‘Romans and Batavians: a regional study in the Dutch eastern river area II’,Berichten
van de Rijksdienst voor het Oudheidkundig Bodemonderzoek 34, 39–331
Willems, W.J.H., van Enckevort, H., and Thijssen J.R.A.M. 2009: ‘From Batavians to Franks: changes in late
antiquity’, in W.J.H. Willems and H. van Enckevort (eds), VLPIA NOVIOMAGVS Roman Nijmegen: The
Batavian Capital at the Imperial Frontier, Journal of Roman Archaeology Supplement 73, Portsmouth, RI,
95–105
Wolters, R. 2001: ‘Germanische Mobilität und römische Ansiedlungspolitik: Voraussetzungen und
Strukturen germanischer Siedlungsbewegungen im römischen Grenzland’, in T. Grünewald (ed.),
Germania inferior: Besiedlung, Gesellschaft und Wirtschaft an der Grenze der römisch-germanischen
Welt, Berlin and New York, 146–68
Woolf, G. 2017: ‘Moving peoples in the early Roman Empire’, in E. Lo Cascio and L.E. Tacoma (eds), The
Impact of Mobility and Migration in the Roman Empire, Impact of Empire 22, Leiden and Boston, 25–41
Zöllner, E. 1970: Geschichte der Franken bis zur Mitte des sechsten Jahrhundert, Munich
NICO ROYMANS ET AL.30