Transformations in North-Western

Europe (AD 300-1000)

Neue Studien zur Sachsenforschung 3

 Transformations in North-Western
Europe (AD 300-1000)
Proceedings of the 60th Sachsensymposion
19.-23. September 2009 Maastricht
Neue Studien zur Sachsenforschung
Band 3

herausgegeben vom
Niedersächsischen Landesmuseum Hannover

in Verbindung mit dem

Internationalen Sachsensymposion

durch
Babette Ludowici
 Transformations in North-Western
Europe (AD 300-1000)
Proceedings of the 60th Sachsensymposion
19.-23. September 2009 Maastricht

herausgegeben von

Titus A.S.M. Panhuysen

Umschlaggestaltung: Karl-Heinz Perschall
Satz und Layout: Karl-Heinz Perschall

Redaktion: Titus A.S.M. Panhuysen, Babette Ludowici

Bibliografische Information
der Deutschen Nationalbibliothek:
Die Deutsche Nationalbibliothek verzeichnet diese
Publikation in der Deutschen Nationalbibliografie;
detaillierte bibliografische Daten sind im Internet über
http://dnb.d-nb.de abrufbar.

© 2011 Niedersächsisches Landesmuseum Hannover Gedruckt mit Unterstützung von:

Alle Rechte vorbehalten
In Kommission bei Konrad Theiss Verlag GmbH, Stuttgart

Abbildungsnachweise
liegen in der Verantwortung der Autoren

Druck:
BWH GmbH - Die Publishing Company, D 30457 Hannover

ISBN 978-3-8062-2576-1
Vorwort
Der vorliegende Band führt die gemeinsam vom Niedersäch-
sischen Landesmuseum Hannover und dem Internationalen
Sachsensymposion herausgegebene Reihe „Neue Studien zur
Sachsenforschung“ fort. Er umfasst die Beiträge zum 60. In-
ternationalen Sachsensymposion, das vom 19. bis 23. Sep-
tember 2009 in Maastricht stattfand. Im Mittelpunkt der
Zusammenkunft stand das Thema „Transformations in North-
Western Europe (AD 300-1000)“. In dieser Zeit vollzog sich in
Europa ein tiefgreifender Wandel: Auf der Grundlage der Kul-
tur der Spätantike formierte sich die Welt des Mittelalters. Die
hier in deutscher, englischer und französischer Sprache vor-
gelegten Konferenzbeiträge gewähren breiten Einblick in die
aktuelle archäologische Forschung zu diesem Thema. Dabei
wird vor allem die enorme Vielschichtigkeit und regionale Va-
riabilität der Jahrhunderte andauernden Transformation im
Nordwesten Europas deutlich. Die 24 Studien zeigen auf, dass
die sehr unterschiedlich geprägten Kulturlandschaften Nord-
westeuropas und ihre Bevölkerung jeweils ganz eigene, kon-
textspezifische Veränderungsprozesse durchliefen.
Für das mehrtägige Symposion waren in Maastricht mehr als
100 Teilnehmer aus Belgien, Dänemark, Deutschland, Frank-
reich, Großbritannien, Niederlande, Norwegen, Schweden und
den USA zusammengetroffen. Die Organisation der Konferenz
wurde von der Universität von Amsterdam und der Stadt Maas-
tricht betreut (Prof. Dr. Frans Theuws und Dr. Titus Panhuysen).
In seinem Festvortrag hat Frans Theuws auf die zentrale Be-
deutung Maastrichts für die frühmittelalterliche Landschaft an
der mittleren Maas hingewiesen, aus der heraus sich in karo-
lingischer Zeit ein neues europäisches Imperium formiert hat.
Fragen nach der Rolle der weltlichen und der klerikalen Elite,
der Bedeutung von Bestattungsritualen sowie der Funktion von
Machtzentren, Siedlungen und des Warenverkehrs in diesem
Prozess waren der Ausgangspunkt von mehreren Forschungs-
projekten der Universität von Amsterdam, so beispielsweise dem
"Sankt Servatius Projekt Maastricht" und dem "Anastasis Grä-
berfeldprojekt der südlichen Niederlande". Am Beispiel ver-
schiedener Studien zu Maastricht, Aachen, Tongeren und Lüttich
konnten im Rahmen der Konferenz aktuelle Erkenntnisse der
Frühgeschichtsforschung in der „Euregio Maas-Rhein“ präsen-
tiert werden. Die Gegenüberstellung dieser Forschungen mit ver-
gleichbaren Vorhaben in anderen Ländern und Landschaften
Nordwest-Europas, die in ganz unterschiedlichen historischen
Entwicklungs- und auch Forschungstraditionen stehen, hat sich
als äußerst ertragreich erwiesen. Eine Reihe internationaler
archäologischer Fundorte von überregionaler Relevanz wie
Spong Hill, Ipswich, Dunum, Krefeld-Gellep, Sorte Muld,
Stavnsager, Uppåkra, Broechem und Bossut-Gottechain bilde-
ten den weiteren Bezugsrahmen. Die Exkursion schließlich
führte die Konferenzteilnehmer zu einigen archäologischen
Spitzenfundorten im belgischen Maastal: Lüttich, Amay, Thier
d'Olne und Namur.
Den Druck des Konferenzbandes haben die Koninklijke Neder-
landse Akademie van Wetenschappen (Königliche Niederlän-
dische Akademie der Wissenschaften) in Amsterdam, die Uni-
versität von Amsterdam und die Gemeinde Maastricht finanziell
unterstützt, wofür wir herzlich danken. Unser Dank gilt außer-
dem unseren Kollegen Barry Ager (London), Diana Briscoe (Lon-
don), Samantha Lucy (Cambridge), Leslie Webster (London) und
Martin Welch (†) für die Durchsicht und sprachliche Betreuung
der englischen Beiträge. Den Bibliographien liegen die in den
Niederlanden üblichen Zitierregeln zu Grunde.
Wir möchten die Aufsatzsammlung dem Andenken an
Martin Welch widmen, der im Februar 2011 verstorben ist.
Sein zusammen mit Sue Harrington verfasster Beitrag zum 60.
Internationalen Sachsensymposion ist seine letzte Veröf-
fentlichung.
Titus A.S.M. Panhuysen
Universität von Amsterdam
Babette Ludowici
Niedersächsisches Landesmuseum Hannover,
Arbeitsbereich „Sachsenforschung“
Claus von Carnap-Bornheim
Leitender Direktor der Stiftung Schleswig-Holsteinische Landesmuseen,
Vorsitzender des Internationalen Sachsensymposions
5
Contents

Frans Theuws
Introduction: Transformations in North-Western Europe (AD 300-1000) 9

Part one: Transformations 11

Menno Dijkstra and Henk Van der Velde

House plots, pots and pins. Transformations in the Rhine estuary during the Early Middle Ages 13

Charlotte Fabech
War and rituals. Changes in rituals and transformations of power 27

Julia Gräf
Die Entwicklung der Gerberei am Übergang von der Spätantike zum Frühen Mittelalter 37

Denis Henrard et Jean-Marc Léotard

Liège au Haut Moyen Âge: un état de la question 47

Annet Nieuwhof
Discontinuity in the Northern-Netherlands coastal area at the end of the Roman Period 55

Titus Panhuysen
Wendepunkte in der Frühgeschichte der Maastrichter Servatiusabtei 67

Alexandra Pesch
Die schwedischen Goldhalskragen. Germanische Bildersprache in antiker Techniktradition. Ein Vorbericht 90

Christoph Reichmann
Die Anfänge des Kirchenbaus im Umfeld des fränkischen Fürstensitzes von Krefeld-Gellep 101

Andreas Schaub
Zur Siedlungskontinuität in Aachen zwischen römischer und karolingischer Zeit 119

Alain Vanderhoeven
Changing urban topography in late Roman and early medieval Tongeren 128

Margrethe Watt
Sorte Muld, Bornholm, an example of transformation and regional contacts during the 5th to 7th centuries 139
in the Baltic Sea area

Sue Harrington and Martin Welch

The archaeological evidence for state formation in southern England: a comparison of the early kingdoms 149
of Kent, Sussex and Wessex

6
Part two: Central Places 159

Reno Fiedel, Karen Høilund Nielsen and Christopher Loveluck

From hamlet, to central place, to manor. Social transformation of the settlement at Stavnsager, eastern Jutland, 161
and its networks, AD 400-1100

Lars Larsson
Power by fire. Transitions and continuity during the Migration and Merovingian periods at Uppåkra, 177
southernmost part of Sweden

Ulf Näsman
Central Places in South Scandinavia – A Transformation Twenty Years After 185

Martin Rundkvist
Transformations in the elite settlements of Östergötland, Sweden, 375-1000 194

Christopher Scull
Ipswich – Transformations of Community and Settlement in the Seventh and Eighth Centuries 200

Part three: Burial Customs 205

Rica Annaert, Koen Deforce and Marit Vandenbruaene

The cremation graves at the Broechem cemetery (prov. of Antwerp, Belgium) 207

Egge Knol
The first early medieval cemeteries along the northern Dutch coasts and their significance for Anglo-Saxon migration 218

Daniel Peters
Am Vorabend der Sachsenkriege: Aufkommen und Abbruch der sogenannten sächsischen Gräber in Westfalen 228
am Beispiel von Soest

Annette Siegmüller
Leichentücher und Federstreuungen. Das frühmittelalterliche Gräberfeld von Dunum als Spiegel politisch-religiöser 239
Wandlungen des 7.-10. Jahrhunderts im Küstenraum

Anne Birgitte Sørensen

Changes in Burial Custom seen from Østergård, Southern Jutland in the period from the Early Roman Iron Age 251
to the Late Roman/Early Germanic Iron Age

Olivier Vrielynck
The Merovingian cemetery of Bossut-Gottechain (Grez-Doiceau, Belgium) 259

Leslie Webster
The Prittlewell (Essex) burial: a comparison with other Anglo-Saxon princely graves 266

Liste der Teilnehmer 273

7
Discontinuity in the Northern-Netherlands coastal area at the end of the Roman Period

Annet Nieuwhof

Figure 1. Palaeogeographical map of the coastal area of the northern Netherlands, AD 800 (after Vos/Knol 2005). Place-names mentioned in the text:
1 Driesumerterp; 2 Dronrijp; 3 Englum; 4 Ezinge; 5 Midlaren; 6 Wijnaldum.

There are indications that inhabitation of the coastal area of
the northern Netherlands (Figure 1) came to an end during
the late Roman Period, and that immigrants from the north-
east (‘Anglo-Saxons’) came to occupy the salt marsh area from
the end of the fourth century AD onwards. This has already
been discussed in various publications.1 However, as yet these
ideas have not reached a wide audience beyond the small cir-
cle of archaeologists who work in the Dutch coastal area.
Moreover, discontinuity is hard to prove from archaeological
evidence, while migration is not a popular model to explain
change in archaeology, so discontinuity and immigration are
still not widely accepted as an explanation of the changes
which occurred in this area at the time. This article means to
rekindle the discussion on the subject. By presenting it to an
international audience, it will hopefully become part of the
discussion and research on the changes that occurred in many
parts of the North Sea coastal area at the end of the Roman
Period and the early Middle Ages, and contribute to a better
understanding of these changes.
The arguments from the discussion thus far will be sum-
marised. The arguments for discontinuity are supported by
new evidence from research of the pottery from the inland set-
tlement of Midlaren-De Bloemert. Thereby, it is possible to
show what continuity in this area might look like, implicitly
showing how we may recognise discontinuity. Finally, the rea-
sons for the break in habitation in this area will be discussed.
Arguments for discontinuity and immigration
The idea of an invasion of new Anglo-Saxon inhabitants had
already been brought forward by one of the leading experts of
early ‘terp’ archaeology, P.C.J.A. Boeles, as early as 1906.2
Boeles did not think that the original Frisian population had
left the area prior to the invasion. He assumed that the origi-
nal population had suffered from an aggressive Anglo-Saxon
invasion and that the remaining population was absorbed by
the immigrants, thus forming a new Anglo-Frisian population.
His ideas were based on the new material culture (in particu-
lar pottery and cruciform brooches) found in mixed cremation
and inhumation cemeteries during commercial levelling of
terps. The new pottery and the cruciform brooches were recog-
nised by him as coming from the lower Elbe region.
Boeles’ ideas were not received well at all in Friesland.3

1 See Gerrets 1995; Gerrets 1996; Taayke 1996; Gerrets/Koning 1999; Taayke 1999; Bazelmans 2000; Bazelmans 2001; Taayke 2003; Koning 2003;
Bos/Brouwer 2005; Nicolay 2005; Nicolay 2006; Taayke 2008; Knol 2009.
2 Terp is the Frisian word for the artificial dwelling mounds of the northern Netherlands; it is the term most often used in international publications by Dutch
authors. The term wierde is used to describe similar mounds in the province of Groningen. In this article, terp will be used for both areas.
3 Bazelmans 2000.

55
Nationalist tendencies there were strongly inclined towards a
direct descent from the Frisians mentioned by the classical au-
thors, who were reported to have resisted the Romans suc-
cessfully. Moreover, there were historians who rejected
immigration as an explanation of changes in material culture
on principle. For example, the historian Slicher van Bath re-
jected the theory that objects could be equated with ethnic
groups; he thought it more likely that the northern-Nether-
lands area, as a part of the North Sea cultural sphere, was in-
fluenced by groups that were dominant in certain periods,
such as the Saxons in the fifth century. This influence had per-
suaded the Frisians to accept their material culture and lan-
guage.4 The argument of Slicher van Bath is, of course, still
persuasive. However, during the 1990s, new evidence was
presented which supported Boeles’ ideas of an Anglo-Saxon
immigration. There appeared to be indications of a break in
habitation well before the arrival of the immigrants. It had not
been a hostile take-over. The new evidence was provided in
particular by the study of indigenous pottery by Taayke (1996)
and by the results of the excavation of Wijnaldum (1991-
1993).
Arguments for discontinuity can be divided between evi-
dence for the abandonment of the area c. AD 300 on the one
hand (nos. 1-6), and for immigration at a later stage on the
other hand (nos. 7-10):
1 There is a general and strong decline in finds from the
middle-Roman Iron Age onwards (e.g. in Englum).5
2 In the salt marsh region, pottery of the last phase (c. AD
250 to 350, so-called Driesum-style pottery) is not found
with pottery of the Migration Period. Moreover, there are
no transitional forms between Driesum-style and Migra-
tion Period pottery found in this area.6
3 Hardly any pottery from the fourth century AD has been
found in the coastal area; however, regional differences
do occur. In the western part of the present province of
Friesland, Westergo, pottery types from the fourth century
are entirely missing. In the eastern part of the province,
Oostergo, Driesum-style pottery is possibly used into the
early fourth century. In Groningen, a small number of pots
of fourth century types known from northern Drenthe and
northwestern Germany have been found.7
4 There are hardly any metal finds known from the fourth
century AD in the salt marsh area.8
4 Slicher van Bath 1949.
5 Nieuwhof 2008a.
6 Taayke 1996-97, 180.
7 Taayke 1996-97, 51-52, 55, Abb. 10.
8 Erdrich 1999, 177.
9 Cf. Bazelmans 2000; Bazelmans 2001.
10 Bos/Brouwer 2005.
11 A dissertation on Iron Age funerary rites in the Dutch coastal region is being prepared by the author.
12 Knol 2009; Knol, this volume.
13 Taayke 1996-97, 195; Gerrets/Koning 1999, 104; Dijkstra et al. 2008, 318.
14 Taayke 1996-97, 163 ff.
15 Schmid 2006, 78.
56
5 There is usually a hiatus of decades (at least) between
the youngest dated Roman Iron Age terp layers and the
oldest layers with an early medieval (including Migration
Period) date.
6 Prehistoric place and river names are entirely absent in
Friesland and virtually absent in Groningen; place and
river names are usually not older than the early Middle
Ages.9
7 Specific brooch types appear in the fifth century AD, in
particular cruciform brooches.10
8 While cemeteries from the pre-Roman Iron Age and the
Roman Period are entirely lacking in the salt marsh area
of the northern Netherlands, and only a small number of
single graves and separate human bones have been
found, formal cemeteries are introduced in the fifth cen-
tury AD, with cremations and inhumations.11 The new
cemeteries are quite similar to cemeteries in the German
Bight and in England.12
9 In the fifth century AD, the orientation of houses and
other structures in settlements change, compared to the
late-Roman Period. Houses with a wooden framework
are no longer found, while sod houses are introduced.13
10 New pottery types of (Anglo-)Saxon style occur in the late
fourth/fifth century in the salt marsh area.
Recognition of continuity
Additional evidence is provided by contemporary develop-
ments in areas where a break in habitation is not assumed, in
particular in northern Drenthe. There, continuity can be
demonstrated from a continuous development of pottery
styles. Since pottery is used as evidence here, some back-
ground information follows.
The study of handmade pottery by Taayke revealed that the
development of pottery styles in the northern Netherlands was
strongly influenced from the east.14 This trend is sometimes
called ‘Germanisation’, suggesting a certain cultural, social or
political influence as well. However, this influence was not a
one-way affair; pottery types from the northern Netherlands
(‘Frisian pottery’) have been found as far east as the Elbe-
Weser area, so ‘Germanisation’ may not be the right word for it.15
In the first century AD, the influence from the east had
spread to Groningen and northern Drenthe, while the Frisian
Figure 2. Midlaren-De Bloemert. A sequence of beakers from the middle-Roman Period (top rows), to the Migration Period (bottom rows). There is a gradual
change in shapes and decoration.

57
and North-Holland areas still had their own style which had
developed from earlier indigenous types.16 From AD 100 to
250, the eastern influence expanded to the south, taking in
the larger part of Drenthe; the pottery of the time in this area
belongs to the so-called nordseeküstennahen Fundgruppe.
Friesland and North-Holland still had their own, different, pot-
tery style. In the third century, new shapes evolved from older
ones in Groningen, Drenthe and northwestern Germany; these
are also found in the eastern part of present Friesland (Oost-
ergo). Taayke called this the ‘Driesum-style’, after one of the
Oostergo locations.17 Driesum-style pottery was also found in
North-Holland and Westergo in several locations.18 However,
finds dating to this period were already becoming scarcer. We
may conclude that the process of ‘Germanisation’ had resulted
in a rather homogeneous pottery style in the third and early
fourth century AD in the inland, as well as in the entire coastal,
areas of the northern Netherlands, in so far as they were still
occupied.
These similarities between the pottery of the coastal area
and the Pleistocene inland make it suitable for a comparison
of the developments in both areas. Taayke argued on the basis
of his sample of northern-Drenthe handmade pottery, that in
northern Drenthe (as well as in northwestern Germany), the
angular forms of the middle-Roman Period pottery gradually
changed into S-shaped profiles.19 The Driesum-style was part
of this development. At the same time, the rather formal deco-
ration on the funnel-shaped beakers of the middle-Roman Pe-
riod gradually came to be replaced by the expressive
decorative elements which we usually call Saxon or Anglo-
Saxon.20
The pottery of the recently excavated settlement of Mid-
laren-De Bloemert, which considerably extended the pottery
sample that was available to Taayke, clearly illustrates this
gradual development.21 A small selection of the Midlaren pot-
tery is depicted in figures 2, 3 and 4.22 Figure 2 shows the de-
velopment of beakers, starting with the middle-Roman Period
in the top rows. Decoration consists of formal triangular en-
gravings and zones with dots (nos. 1918 and 3292) or, some-
what later, of horizontal formal zones (no. 1894, an
exceptionally large beaker). No. 655 is a beaker of the late
third or early fourth century. The lower three rows represent
fourth and fifth century beakers, with and without Anglo-
Saxon style decoration. Figure 3 shows a find assemblage in
Driesum-style, with a decorated sherd (no. 1539) indicating
that it may be an early fourth-century assemblage. Figure 4
shows cooking pots and other vessel-types, starting with some
angular middle-Roman Period rims in the upper left corner,
then another Driesum-style sherd (no. 957), and the more flu-
ent and rounded shapes that follow. Some of these vessels,
such as the striking Buckelurne (no. 2582), may have been
imported from elsewhere and influences from elsewhere may
also be recognised. However, most pots were undoubtedly
made locally and there is nothing to suggest that they were
made by immigrants.23
The evidence not only shows that there is probably no
break in habitation in the northern-Drenthe area between the
Roman Period and the early Middle Ages; it also shows how
continuity might be recognised, thereby implicitly showing
how we may recognise discontinuity. When we compare the
pottery of the settlement of Midlaren-De Bloemert with the
finds from any settlement in the coastal area, it is clear that
there was no continuous development of pottery in the coastal
area. The shapes and decorations of the abundant fourth-cen-
tury pottery of Midlaren, which connect the funnel-shaped
beakers and the Driesum-style pots with distinctive Anglo-
Saxon style pots, are entirely lacking in the coastal area.
A good example is the pottery of Englum, found in this
Groningen terp during the excavation of 2000.24 Here,
Driesum-style pots are the last to occur at the end of the
Roman Period. Figure 5 shows a find assemblage that is quite
similar to the find assemblage from Midlaren shown in figure
3, with a large number of broken Driesum-style pots found to-
gether. The presence of a carinated pot (no. 253) indicates
that the Englum assemblage is from the early Driesum-phase,
not long after the middle of the third century, while the Mid-
laren assemblage may be half a century later. It is notable that
in Midlaren there were dozens of similar pots from other con-
texts, but in Englum there were only five contemporary sherds
besides this assemblage.25 Some Anglo-Saxon-style pottery is
only found in later layers and features.26 It seems to be some-

16 The names of Dutch provinces may be confusing to foreigners. North-Holland is not the northern part of the country, but the most northern province of the
western Netherlands (‘Holland’). The northern Netherlands consist of the provinces of Friesland and Groningen along the coast, and the province of Dren-
the to their south. The early ‘Frisians’ supposedly lived in all these areas, including North-Holland.
17 Taayke 1996-97, 180.
18 Taayke 1996-97, 194.
19 Taayke 1996-97, 180; Taayke 1999, 199.
20 A similar, though not identical, development of pottery was recognised in the Feddersen Wierde material (Schmid 2006).
21 In the Midlaren material, there were fragments of 965 individual pots from the second and third century AD, 310 from types of the fourth century AD and
518 from ‘Anglo-Saxon’-style pottery, c. 350-550 AD (Nieuwhof 2008b, 291).
22 Pottery drawings were made by the author.
23 Nos. 2873, 2874, 2877 and 2878 (Figure 4), as well as no. 2875 (Figure 2) were found associated; so where no. 2931 (Figure 4) and no. 2930 (Figure 2);
and nos. 3608, 3611 and 3612 (Figure 4). Nos. 2581, 2582 (Figure 4) and no. 2583 (Figure 2) were found in a mixed cemetery, as either cremation urns
(2581 and 2582) or grave gifts in an inhumation grave (2583).
24 Nieuwhof 2008a.
25 In Englum, the fragments of 639 individual pots from the first century AD, 75 from the second and third century AD and only 13 from the third and possi-
bly early fourth century were found (Nieuwhof 2008a, 63).
26 N = 12.

58
Figure 3. Midlaren-De Bloemert. A find assemblage consisting of Driesum-style pottery. Sherd no.1539 indicates an early fourth century date for the assemblage.

59
thing entirely new and different in the area. The most con-
spicuous find (Figure 6) is a large fragment from a Schalenurne
found in a well, together with a scabbard slide. The scabbard
slide is a unique find in the Netherlands; its shape and deco-
ration suggest an eastern origin (possibly the Elbe or Weser
area); both finds are dated to the 5th century.27
When and where?
On the basis of this and similar evidence, it can be safely as-
sumed that discontinuity in pottery production is an indica-
tion of discontinuity in the habitation of the coastal area. So
when and where exactly did this break in habitation occur?
The western part of Friesland, Westergo, was almost en-
tirely abandoned. The find of a cremation burial together with
a terra nigra-like pot, presumably from the fourth century, in
Dronrijp28 is sometimes taken to show that a very small popu-
lation remained there.29 However, a recent radiocarbon date
showed that this cremation might as well be dated to the fifth
century AD.30 North-Holland was largely abandoned as well,
with the exception of a small number of settlements where
habitation possibly did not end completely.31 The eastern part
of Friesland, Oostergo, was the most densely populated area
in the terp region in the third century, and again in the fifth
century. However, a fourth-century finds horizon is absent. In
Groningen, too, most terps seem to have been deserted, but
there is evidence that a small population remained on some
terps, for example Ezinge.32 Groningen may not have been so
devoid of people as other parts of the terp region.
The abandonment of the coastal area was a process that
may already have started early in the third century. It was the
remaining and depleted population that adopted the Driesum-
style. However, in the early fourth century, the remaining pop-
ulation left as well. For example in Wijnaldum, it could be
established that habitation ended in the first quarter of the
fourth century.33 In most areas, an end date cannot be estab-
lished so accurately. Nevertheless, there are indications that
over a thousand settlements were abandoned at the end of
the Roman Period.34 New habitation in Wijnaldum started c.
AD 425, a century later. In other areas, it may have started
somewhat earlier, possibly at the end of the fourth century.35
In Germany, the picture is somewhat different. A break in
habitation is assumed for parts of the coastal area of north-
27 I would like to thank Prof. Dr. Claus von Carnap-Bornheim and Dr. Andreas Rau (Schloss Gottorf) for their help in finding information on the scabbard slide.
28 Nieuwhof 2008c.
29 Taayke 1996-97, 195.
30 GrN-31590: 1640 ± 25 BP.
31 Koning 2003.
32 Taayke 1996-97, 195.
33 Gerrets/Koning 1999, 99.
34 Taayke 2005, 200.
35 Taayke 2005.
36 Bärenfänger 2001.
37 Taayke 1996-97, 194; Taayke 1999, 200.
60
western Germany as well, but there are significant regional
differences.36 Two factors complicate comparison between
northwestern Germany and the northern Netherlands. In the
first place, our knowledge of the break in habitation in the
Dutch coastal area is based on the evidence from a small num-
ber of excavations, but supported by thousands of finds from
hundreds of terps which have been destroyed by quarrying the
fertile soil. The German Wurten were not levelled, however,
which limits the evidence to finds from excavations.
In the second place, there is an important difference bet-
ween the landscapes of the northern coastal areas of the
Netherlands and Germany, namely the size of the salt marsh
area and the vicinity and accessibility of the Pleistocene in-
land (in Germany called the Geest). In the Netherlands, the
Holocene terp region was an extensive landscape and the in-
habited parts of the Pleistocene inland were nowhere near
most terp settlements, especially in Friesland. Moreover, it was
not always easy to reach the interior; access was hindered by
the extensive peat zone between the salt marshes and the
Pleistocene inland. In Germany, the salt marshes often formed
a rather limited zone along the coast, with easy access to the
nearby Geest. In the Dutch part of the coastal area, inhabita-
tion of the Holocene salt-marsh landscape came to an end,
while it continued in the Pleistocene inland. For the German
part of the coastal area, such a difference has not yet been
established.
Reasons for leaving
The Dutch situation, where abandonment is related to one
type of landscape only, is rather suggestive as to the causes of
the abandonment of the area. However, it is not yet entirely
clear what actually caused the break in habitation. Various
push-and-pull factors have been mentioned: increased flood-
ing; epidemics; (political) pressure coming from the east; tribal
unrest; an economic crisis caused by the collapse of the
Roman Empire; or the attraction of the Roman Empire when
it collapsed.
Taayke has already suggested that natural causes, in par-
ticular problems with drainage, were the primary reason for
leaving, with tribal unrest caused by the collapse of the Roman
Empire as a secondary reason.37 However, the results of the
Wijnaldum excavation in the western part of Friesland, West-
Figure 4. Midlaren-De Bloemert. Angular, middle-Roman Period shapes (upper row left) and following late-Roman Period and Migration Period pottery.

61
Figure 5. Englum (Prov. Groningen). A find assemblage consisting of Driesum-style pottery. First half of the 3rd century.

ergo, have cast some doubts on these deductions. It was clear
that a high cap ridge had formed in the north of Westergo
which had closed off inland watercourses (this was beautifully
mapped by Vos).38 It was argued that Wijnaldum was situated
on a rather high salt marsh ridge that would not have suffered
from drainage problems. Moreover, there probably were open
creeks near Wijnaldum in the period that the settlement was
abandoned, providing for drainage of the area north of the
settlement. Nevertheless, it was thought that stagnating in-
land water might well have posed major difficulties to the in-
habitants of the lower, older salt marshes, more to the south.39
Vos and Knol still share this opinion and suggest that natural
causes were probably not the prime mover of the break in
habitation.40 However, they do observe that many settlements
were silted over during the Roman Period.41
It is quite certain that the abandonment was not caused by

38 Vos 1999, fig. 23.

39 Gerrets/Koning 1999, 99 ff.; Vos 1999; Bazelmans 2000; Bazelmans 2001.
40 Vos/Knol 2005, 128.
41 Vos/Knol 2005, 126.

62
Figure 6. Englum (Prov. Groningen). Finds from the Migration Period. The potsherd and the scabbard slide were found together in a well (drawing scabbard slide
M.A. Los-Weijns, University of Groningen).
increased marine flooding. The terp dwellers had been used to
marine flooding for centuries; in fact, living on an artificial
dwelling mound was a deliberate and adequate adaptation
to life in a landscape that was occasionally flooded by sea-
water. Flooding could be handled as long as it did not last
much longer than a few days and the area was well drained.
However, the sea level rise that continued while the salt
marshes expanded to the north over the centuries, not only
resulted in an ever larger salt marsh area, but also in an ever
higher area (a cap ridge) along the northern coastline.42 This
caused drainage problems inland, with prolonged periods of
flooding during which thick layers of clay were deposited. This
may have made life in these areas quite difficult. Drainage
problems were only solved later, when new tidal systems and
watercourses started to drain the inland areas. Their dating is
as yet not well established.
Thick layers of clay were deposited in the oldest, low areas
of the salt marsh across the entire coastal area. However, these
are not all dated to exactly the same period (as far as they are
dated). Still, older and more recent excavations in coastal set-
tlements in Friesland, as well as in Groningen, have provided
evidence of stagnating water and of prolonged periods of
flooding from the middle-Roman Period onwards; this coin-
cided with the abandonment of many settlements in the third
century.43
Without wanting to be too ecologically deterministic, it
cannot be denied that environmental causes may have played
an important part in the end of habitation. They may well have
42 = The high salt marsh ridges in Figure 1.
43 For example, the ‘frustrated terps’ of Paddepoel in Groningen (Es 1970); Hoxwier in Westergo (Nieuwhof/Prummel 2007).
44 Knottnerus 1999; 2002.
been the prime mover, though dating is not yet well estab-
lished everywhere. The argument that settlements on high salt
marsh ridges, such as Wijnaldum, were not affected by
drainage problems, may not be so strong. If low parts of the
surrounding salt marsh were flooded for long periods, it would
affect life in many ways. Moreover, malaria mosquitoes may
have thrived in the stagnating brackish water, thus contribu-
ting to an increasingly unhealthy environment.44 A high death
rate caused by malaria would have weakened the population
considerably. With or without malaria, as soon as the first set-
tlements were abandoned and groups of people started to
leave the area, social networks weakened; they may finally
have collapsed. For the people staying behind, there was no
longer a reason to stay, whether they lived on high ridges or
not.
Such a process also makes it understandable why habita-
tion of the Frisian areas ended almost completely, while in
Groningen a small number of terp settlements remained in
use. Although natural conditions in Groningen did not differ
much from those in Friesland, the social network was much
stronger, because it included areas outside the Groningen terp
area where natural conditions did not deteriorate. Since the
early-Roman Period, the inhabitants of Groningen had been
part of a social network, which included northern Drenthe and
northwestern Germany. Habitation in these areas did not come
to an end during this period, so only a part of the territory of
this group (that shared cultural and social identities) became
uninhabitable. Habitation of terps in well-drained areas could
63
continue, because a large part of the social and cultural net-
work of the inhabitants had remained intact.
In Friesland, the situation was quite different. The popula-
tion of Friesland, especially in Westergo, had been more ori-
ented to the west than to the east, and was not part of a social
network which largely remained intact while the terp region
was abandoned. On the contrary, the region to which the in-
habitants of Friesland were primarily connected, North-Hol-
land, was also abandoned in this period. It is assumed that the
reasons for the end of habitation there were the worsening
political-military situation, the economic decline and the de-
terioration of the natural environment as a result of overcrop-
ping. These causes were directly related to the collapse of the
Roman Empire at the end of the third century.45 North-Hol-
land had always maintained rather close relations with the
Roman Empire, probably more so than the northern coastal
area.46
It is not clear where the terp inhabitants went after they left
the area. Traces of them, such as an increase in the number
and size of settlements, have not been observed elsewhere.
Driesum-style pottery, the pottery of a large part of the emi-
grants, has not been found in other parts of the Netherlands.47
However, pottery in this style was discovered some years ago
in Zele in Belgian Flanders, in a deposit that was remarkably
similar to the find assemblages from Englum and Midlaren
that were illustrated in figures 3 and 5.48 It has been suggested
that the northern Netherlands were among the source areas
of Frankish migrants, who moved south during the fourth cen-
tury.49 Though this may well be so, it is possible that part of the
population, especially in Groningen, did not join them but
stayed within their cultural and social environment, only mov-
ing a relatively short distance to the east or to the south. Since
they shared the same material culture, the newcomers may
not be recognisable as separate groups in either northern
Drenthe or northwestern Germany.
New settlers, probably profiting from the improved
drainage of the area after new water courses had developed,
came from the east. This can be established from their mate-
rial culture. In principle, it is possible that the area was re-
populated from the Pleistocene inland which, as we have seen,
developed its own Saxon-style pottery. However, the many im-
ported objects that are not related to the material culture of
Drenthe – for example the scabbard slide mentioned above,
the new cemeteries, the new sod houses and the different set-
tlement structure – make it more likely that they came from the
eastern North Sea area. The region between Weser and Elbe,
45 Bazelmans et al. 2004.
46 Erdrich 2001.
47 Taayke 1996, 195-196.
48 Clerq/Taayke 2004.
49 Taayke 1999, 195; Clerq/Taayke 2004.
50 Cf. Nicolay 2005.
51 Taayke 2008, 203.
52 Cf. Bazelmans 2000; Bazelmans 2001; Bazelmans 2009.
53 See note 52.
64
and Schleswig-Holstein, are both possible areas of origin.50
The new inhabitants are only represented by a relatively small
number of finds compared to the remains of habitation from
the Roman period, and new terps were only started in the se-
venth century.51 The population probably did not grow rapidly
in the early years, and only relatively small groups of settlers
may have come to the area at first.
What’s in a name?
Despite all the evidence for a break in habitation and the ar-
rival of immigrants from elsewhere, there is one important ar-
gument for continuity in the coastal area: the continuity of the
name of the Frisians. It was used in historical sources from the
first century, and it is also mentioned in early medieval sources.
How is it possible that the name remained in use, while the
population changed?
There are several possible answers to this problem.52 Firstly,
a small population may have remained so that the newcom-
ers could adopt their name. However, the area of the present
province of Friesland in particular was probably devoid of peo-
ple in the fourth century, while we do not know whether the
small remaining population of Groningen called themselves
Frisians.
Secondly, it is possible that newcomers took the name of
the region they colonised. In that case, the name of the area
of the Frisians existed independently of its inhabitants. It must
have been well-known among other groups living in the North
Sea area, among them the people that came to occupy the
Frisian land. As seafaring people, they may have been famil-
iar with the geography and the geographical names of the en-
tire coastal area of the North Sea.
Thirdly, the new inhabitants may first have been called
Frisians by outsiders, in particular by those of the Frankish elite
who wrote about them. They knew the geography and history
of northwestern Europe from antique written sources and used
the old name ‘Frisians’ to describe the new inhabitants of the
Frisian area. This name was then also adopted by the new
Frisians themselves. This explanation was preferred by Bazel-
mans.53
I would like to add a fourth possibility. If groups of emi-
grants, who called themselves Frisians, indeed settled in north-
western Germany, they may have kept the name and the
memory of their descent alive. If their descendants were
among the new settlers of the northern Netherlands coastal
area, they may have reintroduced the name of the Frisians to
the area.
Although these explanations have their merits, it is not pos-
sible to make a well-founded choice between them. Evidence
is simply too scarce. Nevertheless, it does not seem so strange
that the name of the Frisians continued, though possibly used
for and by a different population. Continuity of a name does
not necessarily exclude a break in habitation.
Conclusion
A possible break in the occupation of the salt marsh area of
the northern Netherlands has long been a subject for debate
in Dutch ‘terp’ archaeology. Many arguments have been
brought forward to support the theory that the area was aban-
doned in the late-Roman Period and repopulated in the late
fourth or early fifth century. A comparison of the development
in pottery styles in an inland settlement, Midlaren-De Bloe-
mert in northern Drenthe, and a salt marsh settlement, En-
glum in Groningen, confirmed the earlier arguments. It is very
clear that a continuous development of pottery styles, which
is apparent from the Midlaren material, is completely absent
in Englum and other salt marsh settlements. There is nothing
between the (already sparse) third century pottery and the
seemingly exotic Anglo-Saxon pottery of the immigrants in the
coastal area.
The prime motive for leaving the area probably was the in-
creasingly problematic drainage caused by the formation of
high cap ridges along the coast. This resulted in prolonged
periods of flooding which must have made life in the area
quite difficult. It is suspected that malaria in this wet, brack-
ish environment thrived, thus adding to the increasing unat-
tractiveness of the area. Many settlements were already
abandoned in the early third century; social networks weak-
ened and finally collapsed. There was no longer any reason
to stay, even for those who lived in well-drained areas. While
the province of Friesland was probably abandoned completely,
a small population may have stayed behind in the Groningen
area. This difference may be explained by social and cultural
differences between these areas. Friesland had been oriented
to the west, to North-Holland, which was also abandoned dur-
ing this period. Groningen had been part of a social and cul-
tural network that included northern Drenthe and reached far
into Niedersachsen. There, the result of staying behind would
not have been total isolation, natural circumstances permit-
ting. Thus, social circumstances may have been the decisive
factor in the end.
Bibliography
Bärenfanger, R., 2001: Befunde einer frühmittelalterlichen Siedlung bei Esens,
Ldkr. Wittmund (Ostfriesland), Probleme der Küstenforschung im südlichen
Nordseegebiet 27, 249-300.
Bazelmans, J., 2000: Een laat-Romeins bewoningshiaat in het Nederlandse
kustgebied en het voortbestaan van de Friezennaam, Jaarverslagen van de
Vereniging voor Terpenonderzoek 76-82, 14-75.
Bazelmans, J., 2001: Die spätrömerzeitliche Besiedlungslücke im niederlän-
dischen Küstengebied und das Fortbestehen des Friesennamens, Emder Jahr-
buch für historische Landeskunde Ostfrieslands 81, 7-61.
Bazelmans, J., 2009: The early-medieval use of ethnic names from classical an-
tiquity. The case of the Frisians, in: T. Derks/N. Roymans (eds.), 2009: Ethnic
constructs in Antiquity. The role of power and tradition, Amsterdam (Amster-
dam Archaeological Studies 13), 321-337.
Bazelmans, J./ M.F.P. Dijkstra/J. de Koning, 2004: Holland during the first mil-
lennium, in: M. Lodewijckx (ed.), 2004: Bruc ealles well. Archaeological es-
says concerning the peoples of north-west Europe in the first millennium AD,
Leuven (Acta archaeologica Lovaniensia monographiae 15), 3-36.
Boeles, P.C.J.A., 1906: De opgravingen in de terp Hoogebeintum, De Vrije
Fries 20, 391-430.
Bos, J.M./E.W. Brouwer, 2005: De kruisvormige fibulae van Friesland, De Vrije
Fries 85, 9-36.
Clerq, W. De/E. Taayke, 2004: Handgemachte Keramik der späten Kaiserzeit
und des frühen Mittelalters in Flandern (Belgien). Das Beispiel der Funde
friesischer Keramik in Zele (O-Flandern), in: M. Lodewijckx (ed.), 2004: Bruc
ealles well. Archaeological essays concerning the peoples of north-west Eu-
rope in the first millennium AD, Leuven (Acta archaeologica Lovaniensia
monographiae 15), 57-71.
Dijkstra, J./D.A. Gerrets/J.A.W. Nicolay, 2008: Synthesis. Origin and develop-
ment of the Oldehove terp at Leeuwarden, in: J. Dijkstra/J.A.W. Nicolay (eds.),
2008: Een terp op de schop. Archeologisch onderzoek op het Oldehoofsterk-
erkhof te Leeuwarden, Amersfoort (ADC Monografie 3), 307-342.
Erdrich, M., 1999: Continuity or discontinuity. Native and Roman metal finds,
in: J.C. Besteman/J.M. Bos/D.A. Gerrets/H.A. Heidinga/J.de Koning (eds.),
1999: The excavations at Wijnaldum, Rotterdam/Brookfield (Reports on Frisia
in Roman and Medieval times 1), 171-184.
Erdrich, M., 2001: Rom und die Barbaren. Das Verhältnis zwischen dem Im-
perium Romanum und den germanischen Stämmen vor seiner Nordwest-
grenze von der späten römischen Republik bis zum Gallischen Sonderreich,
Mainz am Rhein (Römisch-Germanische Forschungen 58).
Es, W.A. van, 1968 [1970]: Paddepoel, excavations of frustrated terps, 200
BC - 250 AD, Palaeohistoria 14, 187-352.
Gerrets, D.A., 1995: The Anglo-Frisian relationship seen from an archaeo-
logical point of view, in: V.F. Faltings/A.G.H. Walker/O. Wilts (eds.), 1995:
Friesische Studien 2. Beiträge des Föhrer Symposiums zur Friesischen Philo-
logie vom 7.-8. April 1994, Odense, 119-128.
Gerrets, D.A., 1996: Continuity and change in house construction and the
lay-out of rural settlements during the early middle ages in the Netherlands,
Ruralia 1, Prague (Památky archeologické-supplementum 5), 33-46.
65
Gerrets, D.A./J. de Koning, 1999: Settlement development on the Wijnaldum-
Tjistma terp, in: J.C. Besteman/J.M. Bos/D.A. Gerrets/H.A. Heidinga/J. de Kon-
ing (eds.), 1999: The excavations at Wijnaldum, Rotterdam/Brookfield (Reports
on Frisia in Roman and Medieval times 1), 73-124.
Knol, E., 2009: Anglo-Saxon migration reflected in cemeteries in the North-
ern Netherlands, in: D. Quast (ed.), 2009: Foreigners in early medieval Eu-
rope. Thirteen international studies on early medieval mobility, Mainz
(Monographien des Römisch-Germanischen Zentralmuseums 78), 113-129.
Knottnerus, O.S., 1999: Malaria in de Nordseemarschen. Gedanken über
Menschen und Umwelt, in: M. Jakubowski-Tiessen/K.-J. Lorenzen-Schmidt
(eds.), 1999: Dünger und Dynamit. Beiträge zur Umweltgeschichte Schles-
wig-Holsteins und Dänemarks, Neumünster (Studien zur Wirtschafts- und So-
zialgeschichte Schleswig-Holsteins 31), 25-39.
Knottnerus, O.S., 2002: Malaria around the North Sea: a survey, in: G.
Wefer/W. Berger/K.E. Behre/E. Jansen (eds.), 2002: Climate development and
history of the North Atlantic realm, Berlin/Heidelberg, 339-353.
Taayke, E., 1996: Die einheimische Keramik der nördlichen Niederlande. 600
v.Chr. bis 300 n. Chr. (PhD thesis University of Groningen), Groningen (=
Taayke 1990; Taayke 1995; Taayke 1996-1997).
Taayke, E., 1999: The Smell of Higher Nectar, in: H. Sarfatij/W.J.H. Verwers/P.J.
Woltering (eds.), 1999: In Discussion with the Past. Archaeological studies
presented to W.A. van Es, Zwolle, 195-204.
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/H.R. Reinders (eds.), 2003:
Essays on the early Franks, Groningen (Groningen Archaeological Studies 1),
1-23.
Taayke, E., 2005: Het noordelijk kustgebied in de ijzertijd en Romeinse tijd,
in: E. Knol/A.C. Bardet/W. Prummel (eds.), 2005: Professor van Giffen en het
geheim van de wierden, Veendam/Groningen, 153-165.
Taayke, E., 2008: In het spoor van Wijnaldum. De Vrije Fries als eigentijds
platform voor archeologische interpretaties, De Vrije Fries 88, 189-211.

Koning, J. de, 2003: Why did they leave? Why did they stay? On continuity ver-
sus discontinuity from Roman times to the Early Middle Ages in the western
coastal area of the Netherlands, in: T. Grünewald/S. Seibel (eds.), 2003: Kon-
tinuität und Diskontinuität. Germania inferior am Beginn und am Ende der rö-
mischen Herrschaft, Berlin/New York, 53-82.
Nicolay, J.A.W., 2005: Nieuwe bewoners van het terpengebied en hun rol bij
de opkomst van Fries koningschap, De Vrije Fries 85, 37-103.
Vos, P.C., 1999: The Subatlantic evolution of the coastal area around the Wij-
naldum-Tjitsma terp, in: J.C. Besteman/J.M. Bos/D.A. Gerrets/H.A. Heidinga/J.
de Koning (eds.), 1999: The excavations at Wijnaldum, Rotterdam/Brookfield
(Reports on Frisia in Roman and Medieval times 1), 33-72.
Vos, P.C./E. Knol, 2005: Wierden ontstaan in een dynamisch getijdenland-
schap, in: E. Knol/A.C. Bardet/W. Prummel (eds.), 2005: Professor van Giffen
en het geheim van de wierden, Veendam/Groningen, 118-135.

Nicolay, J.A.W., 2006: Een koninklijk machtscentrum in vroegmiddeleeuws

Friesland? De interpretatie van goudvondsten uit de late zesde en vroege
zevende eeuw na Chr., De Vrije Fries 86, 33-94. Annet Nieuwhof
University of Groningen
Nieuwhof, A., 2008a: Aardewerk, in: A. Nieuwhof (ed.), 2008: De Leege Wier Poststraat 6
van Englum. Archeologisch onderzoek in het Reitdiepgebied, Groningen
(Jaarverslagen van de Vereniging voor Terpenonderzoek 91), 47-75.
NL 9712 ER Groningen
a.nieuwhof@rug.nl
Nieuwhof, A., 2008b: Het handgemaakte aardewerk, ijzertijd tot vroege mid-
deleeuwen, in: J.A.W. Nicolay (ed.), 2008: Opgravingen bij Midlaren. 5000
jaar wonen tussen Hondsrug en Hunzedal, Groningen (Groningen Archaeo-
logical Studies 7), 261-304.

Nieuwhof, A., 2008c: Crematieresten uit Dronrijp-Noord, in: K. Huisman/K.

Bekkema/J.M. Bos/H.de Jong/E. Kramer/R. Salverda (eds.), 2008: Diggelgoud.
25 jaar Argeologysk Wurkferbân: archeologisch onderzoek in Fryslân,
Leeuwarden, 127-131.

Nieuwhof, A./W. Prummel, 1999-2006 [2007]: Terpsporen in Hoxwier (ge-

meente Littenseradiel, prov. Friesland), Jaarverslagen van de Vereniging voor
Terpenonderzoek 83-90, 9-41.

Schmid, P., 2006: Die Keramikfunde der Grabung Feddersen Wierde (1. Jh. v.
bis 5 Jh. n. Chr.), Oldenburg (Probleme der Küstenforschung 29 = Feddersen
Wierde 5).

Slicher van Bath, B.H., 1949: Dutch tribal problems, Speculum 24, 319-338.

Taayke, E., 1990: Die einheimische Keramik der nördlichen Niederlande. 600
v. Chr. bis 300 n. Chr., Berichten van de Rijksdienst voor het Oudheidkundig
Bodemonderzoek 40, 101-222.

Taayke, E., 1995: Die einheimische Keramik der nördlichen Niederlande. 600
v. Chr. bis 300 n. Chr., Berichten van de Rijksdienst voor het Oudheidkundig
Bodemonderzoek 41, 9-102.

Taayke, E., 1996-1997: Die einheimische Keramik der nördlichen Niederlande.

600 v. Chr. bis 300 n. Chr., Berichten van de Rijksdienst voor het Oudheid-
kundig Bodemonderzoek 42, 9-208.

66