De 14C-chronologie van de Nederlandse Pre- en Protohistorie VI: Romeinse tijd en Merovingische periode, deel A: historische bronnen en chronologische thema's


  • J.N. Lanting
  • J. van der Plicht


Lack of time, but especially limitations regarding the size of papers in this volume forced us to divide our last contribution on chronology into two parts. This part, A, contains a survey of the historical sources and summaries of the major chronological schemes. Part B will appear in Palaeohistoria 53/54 (2011/12) and will contain a summary of the archaeology of the Netherlands during the Roman and Merovingian periods, and a list of radiocarbon dates.

After some introductory remarks in chapter 1, this paper deals with the historical sources in chapter 2. Special attention is given to the recent publication by Scharf (2005) who convincingly demonstrates the Notitia Dignitatum Occidentisto be a present to emperor Iohannes, produced in AD 423, and not a partly out-of-date state almanack. Consequently, the Roman military and civil administration of Britannia was still intact in that year. On archaeological grounds this had already been shown by Böhme (1986b). That the province Germania II still existed in AD 418 is demonstrated by the letter of emperors Honorius and Theodosius II concerning the creation of the diocese Septem provinciae and yearly conventions of representatives of all 17 provinces in Arles (Weidemann, 1980). The Notitia Dignitatum Occidentis mentions civil administrators of Germania prima and Germania secunda, but no military commander in Germania secunda. Scharf(2005) points out that this omission is due to haste and sloppiness. Germania prima had been divided in 422 or 423 into two military districts: the Tractus Argentorantensis and the district of the Dux Mogontiacensis, and was no longer controlled by a Dux Germania prima. Therefore, the Dux Germania prima mentioned in chapter I of the Notitia should be the Dux Germania secunda. The missing chapter XXXIX (probably not produced in time) would have contained the list of troops under the command of this Dux.

Chapter 3 deals with the areas occupied by the historically known tribes. The area occupied by the Batavians became largely depopulated around AD 250 after this tribe lost its special status in AD 212 and its economy collapsed. Usurper Postumus, ruler of the Gallic empire, allowed Franci (i.e. Frisii and Chamavi) to settle in former Batavian territory after AD 260. When these Franci were forced out of the area by Constantius Chlorus in AD 293 the prisoners of war were partly compelled to join the Roman army (the later crack units Batavi and Equites Batavi), partly resettled as serfs on stateowned estates in present-day northern France and Belgium (the Laeti Batavi). In 341/2 emperor Constans seems to have allowed other Frankish intruders (the Salii) to settle in former Batavian territory. The name Salii appears to have been adopted for the occasion. It means just 'companions' (Springer, 1996; 1997). The Lex Salica has nothing to do with these Salii; its name means 'common law', not 'law of the Salii'. In the latter case the name would have been Lex Saliorum.

The Saxons first appear in contemporary sources in AD 356. The name must have been given by the Romans to seaborne marauders on the British and Gallic coasts after their favourite weapon, the sax. Springer (2003; 2005) has shown that Saxons were never a tribe living on the west coast of Schleswig-Holstein. That misconception is based on the Geographica of Ptolemaeus (II, 11, 11 and 13), where originally ABIONEC (Tacitus' Aviones of Germania 40) must have been mentioned. This was subsequently copied as AΞONEC, translated as Axones and later 'improved' to Saxones, because that name was well known by then. The Germanic invaders of Britannia were largely Angles who therefore called their new homeland England. The name Saxones may have been used for German auxilia from north-west Germany in the late Roman army in Britannia. That may explain why the Anglo-Saxon monk Beda in the 8th century was unable to choose between Anglian or Saxon origin for the leaders of the first group of auxilia employed by the British leaders. For Beda the area of origin of the Antiqui Saxones stretched as far south as the Lippe, whereas Angulus, the area of origin of the Angles, seems to have been restricted to the coastal zones of Lower Saxony and Schleswig-Holstein (and not presentday Angeln on the east-coast of Schleswig-Holstein). The Adventus Saxonum is dated by an entry in the Gallic Chronicle of 452 to AD 441/2 (Muhlberger, 1983; Burgess, 1990). This date is confirmed by archaeological evidence.

Until about AD 350 the Frisii occupied the coastal areas between the mouths of the rivers Rhine and Ems. After that date the population of Holland and the western part of the province of Friesland dwindled considerably. In the eastern part of Friesland numbers dropped as well, whereas in Groningen and northern Drenthe population remained more or less stable. Re-population of Friesland and Noord-Holland started around AD 440, and after AD 455 (?) Frisians also occupied the coastal areas of Zuid-Holland south of the Rhine and of Zeeland. In the second half of the 7th century they occupied the still largely unpopulated coastal areas of the Lower Saxony. The Lex Frisionum of 803/4 mentions Frisians living between the Zwin in Zeeland and the mouth of the river Weser.

Chapter 4 deals with provincial Roman chronology in northern Gaul, and especially with its extension into the late 5th century. Of crucial importance is the short Forschungsbericht by Böhme (1987) in which he introduces the Fundgruppen A and B into which the grave goods of German soldiers and their wives in the western part of the empire can be divided. This Forschungsbericht replaces Böhme's monograph of 1974 and re-dates most of the material Böhme dealt with in1974. Strangely enough this Forschungsbericht has not got the attention it deserves, although the new dates have serious implications for chronology elsewhere, as well. Later Böhme (1994) dealt with the phase of the swords of type Krefeld and the sword-belt buckles of type Krefeld-Gellep and other products of late-Roman fabricae in the period c. AD 465-485. These products were formerly seen as early-Frankish/Alamannic.

Other late-Roman products have also been shown to be produced much later than previously thought. According to Hübener (1968), Rädchensigillata ceased to be made around AD 425. Nowadays the youngest examples of this type of pottery are dated to c. AD 525. Similar re-dating has taken place with other types of pottery and with glass.

In chapter 5 the chronology of the Merovingian and early-Carolingian cemeteries is dealt with. Important is that Böhme's Fundgruppen A and B and his phase of the swords of Krefeld-type c.a. (see above) have been incorporated in the late Roman-Merovingian chronology of the German Lower Rhine area by Aouni (1998) and in the Rheinland-chronology of Nieveler & Siegmund (1999). Böhme's Fundgruppe A corresponds with Rheinland-stage 1, Fundgruppe B with stage 2a, and the phase of the Krefeld swords with stage 2b, That means that the dates given to both Fundgruppen and to the phase of the Krefeld swords by Böhme (1987; 1994) are valid.

Chapter 6 deals with relative and absolute chronology in Germania libera during the Roman and early-Merovingian period. Until now an updated version of the chronological system introduced by Eggers (1951; 1955), but based on work by O. Tischler (1880), Blume (1912) and Plettke (1921), has been used. This system comprises the periods B1 (AD 0-70), B2 (AD 70-160/180), C1 (AD 160/180-250/260), C2 (AD 250/60-310/320) and C3 (AD 310/320-375). Thanks to Bemmann (1993) this Eggers-chronology can be extended with phases D1 and D2, that correspond with zones 3 and 4 in the cemetery of Perdöhl in Mecklenburg. Zone 3 is characterized a.o. by brooches of type Nydam, and zone 4 a.o. by early cruciform brooches. Bemmann synchronises D1 with Zeitstufe II, but seems to be unaware of Böhme's renouncement of these Zeitstufen in 1987, and replacement by the Fundgruppen A and B. Brooches of type Nydam can be shown to be contemporary with Fundgruppe A, and the earliest cruciform brooches with Fundgruppe B. Therefore D1 can be dated to c. AD 390-435, and D2 to c. AD 435-465. That means that C3 must have ended around AD c. 390 instead of AD c. 375. Probably the C2/C3 transition of AD 310/320 will have to be re-dated, as well. In section 6.2.3 the differences between Böhme's Zeitstufen I-III of 1974 and Fundgruppen A and B of 1987 are analysed. In 1974 Böhme combined material found in Gallia and in Germania libera, in 1987 he only used material found in the western part of the Roman empire. As a result, a number of objects included in 1974 is no longer represented in 1987 (for example equal-armed Kerbschnitt brooches). On the other hand, the Forschungsbericht of 1987 includes glass objects and bone combs which were largely ignored in 1974 but turned out to be of chronological importance and which occur both in male and female graves. Core elements in the Fundgruppen are the military belts: Kerbschnitt belts being characteristic for Fundgruppe A and simple belts for Fundgruppe B.

A fair number of types included in the Zeitstufen of 1974 return in the Fundgruppen of 1987, directly or indirectly. A number of brooches characteristic of female burials in north-west Germany can be added thanks to the analysis of the Liebenau cemetery by Brieske & Schlicksbier (2005). But their chronology has to be corrected using Böhme's Forschungsbericht of 1987. Liebenau started around AD 425/430, not around AD 390.

Section 6.3 deals with the dating of decorated 'Saxon' pottery and of undecorated handmade pottery of the Hessens-Schortens family. In north-west Germany the most recent publication on this subject is the one by Schmid (2006) on the pottery from Feddersen Wierde, a dwelling mound in the Elbe-Weser triangle. This settlement, however, only provides us with a rough outline of the development of the pottery in this area between 56 BC - c. AD 500 due to both the way the mound was excavated, and the small amount of datable metal objects in well-defined contexts (Schuster, 2006). In fact, Schmid depends heavily on associations of ceramics and metal objects in cemeteries in the same area, largely excavated by Schön, to produce a typochronology for his pots. An additional problem is that both Schmid and Schuster in the case of Feddersen Wierde, and Schön in the case of the cemeteries, ignore Böhme's revised chronology of 1987 and still refer to Böhme's Zeitstufen of 1974. With these revisions it becomes likely that pots of Plettke's type A6 were produced from c. 350 on and pots of types A6a and A7a from the late 4th century on. These types were produced until at least the early 6th century. Types A7α and β (Buckelware) and C cannot be dated because of an insufficient number of associations. Other finds show, however, that these earlier versions of type C started around AD 350 and the later ones were still produced in the first half of the 6th century. The first Buckelware may have been produced in the late 4th century, the latest was produced in the first half of the 6th century.

The dating by Myres (1969) of the earliest 'Saxon' pottery in England can no longer be accepted. According to Scharf (2005), the Notitia Dignitatum Occidentis was produced in AD 423. At that moment Britannia was still part of the Roman empire. Böhme (1986b, combined with his revised chronology of 1987) had already demonstrated by archaeological means that the Roman army was still present in Britannia around AD 435/440. It seems likely, therefore, that the Adventus Saxonum took place in AD 441/2, as is stated in the Gallic Chronicle of 452. The brooches of north-west German type in early female 'Saxon' graves in Britannia are in accordance with that date. But the earliest 'Saxon' pottery in England is of no importance to the dating of early 'Saxon' pottery on the Continent.

In the Netherlands archaeological dating of decorated 'Saxon' pottery is not possible due to absence of sufficient associations with datable metal objects. Instead, radiocarbon dating of charcoal from settlement contexts and of cremated bones from decorated 'Saxon' pots used as urns provided answers. This dating programme was also used to quash of the misconception of a dual origin of decorated 'Saxon' pottery in the Netherlands: in the northern part of the country introduced by invading 'Saxons' from north-west Germany in the 5th century, in the west and centre of the country by contacts with Anglo-Saxon England in the 6th/7th century. The radiocarbon dates show that decorated 'Saxon' pottery was produced in the north-east Netherlands from c. AD 375 on in Drenthe and in Groningen, and do not contradict the assumption of an introduction in Friesland and probably Noord-Holland after c. AD 440, and in Zuid-Holland south of the Rhine after c. AD 455. We postulate that the introduction of decorated 'Saxon' pottery was just another stage in north-west-German influence on pottery production in the Frisian area, starting with Wierum-style ceramics in Groningen and Drenthe at the beginning of our era, and with Driesum-style ceramics in those areas and in Friesland and Noord-Holland from c. AD 200 onwards. When Frisians from Groningen and Drenthe started to reoccupy former Frisian territory in Friesland and Noord-Holland after AD 440, and occupied former Roman territory south of the Rhine in the coastal areas of Zuid-Holland and Zeeland after AD 455 (?), they introduced decorated 'Saxon' pottery and the undecorated handmade pottery of Hessens-Schortens type. In the Netherlands 'Saxon' pottery refers to late-Frisian pottery.

The radiocarbon dates for decorated 'Saxon' pottery and for undecorated handmade pottery of Hessens-Schortens type seem to indicate that undecorated pottery was only introduced to burial ritual when the decorated 'Saxon' pottery went out of use around AD 525. In settlement contexts this undecorated pottery was already in use in the first half of the 5th century.

Section 6.4 deals with the problem of the Rhein-Weser-Germanic pottery. More research is needed, but almost certainly Rhein-Weser-Germanic pottery was still produced in the first half of the 5th century.