De 14C-chronologie van de Nederlandse Pre- en Protohistorie VI: Romeinse tijd en Merovingische periode, Deel B: aanvullingen, toelichtingen en 14C-dateringen
This is the second part of our paper on the chronology of the Roman and Merovingian periods in the Netherlands. The first part appeared in Palaeohistoria 51/52 (2009/10). Contrary to what we stated in part A, this second part does not contain a summary of developments in material culture, grave ritual, settlement structure and house types. The main reason is the large number of important settlement excavations not yet published. A second reason is that we do not want to get involved in further discussions about house types and ‘house typology’, until more complete or almost complete house plans, excavated under ideal conditions in single trenches, identified in the field and well-dated by finds, dendrodates and/or 14C, have become available.
After some introductory remarks in chapter 1, this paper deals in chapter 2 with additions to part A, in part necessary because we failed to spot relevant information, and in part because of new publications.
Chapter 3 deals with the reliability of radiocarbon dates on cremated bone, with the margins of uncertainty of dendrodates of samples with incomplete sapwood, and with wiggle match-dating of wood samples with relatively small numbers of annual rings.
Although laboratory tests and the cremation of a pig on a pyre in Scotland in 2004 have shown that exchange of CO2 between structural carbonate in bone and the atmosphere of the pyre may take place, this seems to be of little importance. In practice cremated bone has turned out to be a very reliable material for radiocarbon dating.
Because we mention dendrodates as well in chapter 4, some attention had to be given to the margins of uncertainty of dendrodates of samples with incomplete sapwood. Till recently, in the Netherlands the sapwood statistics of Hollstein (1965) for West-German oakwood have been used. A survey of work elsewhere in Europe shows that Hollstein’s numbers are likely to be valid in the Netherlands. But archaeologists should be(come) aware of the statistical meaning of the margins of ±5, ±6 and ±8 proposed by Hollstein. These are standard deviations with a certainty of 68%. Due to the skewed frequency curves of numbers of sapwood rings 95% certainty cannot be achieved by doubling the 1-sigma value. At 95% the margins will be asymmetric, like 16 -9/+13 or 23 -11/+18 calculated by Hillam et al. (1981) for West-German oaks younger, resp. older than 100 years. Sapwood statistics proposed by Jansma in 2007 and based on Dutch archaeological wood will not be used in this paper.
Wiggle match-dating has become fashionable in archaeology. We warn against wiggle match-dating samples with a limited number of annual rings, but in case one does we advise to use the wiggle match-option of OxCal. This programme calculates standard deviations, contrary to the Groningen programme, and these standard deviations show whether wiggle matching is meaningful.
Chapter 4 consists of lists of radiocarbon dates for Roman and Merovingian periods by province. Published dates produced outside Groningen have been included, as are dendrodates of sites with radiocarbon dates. In a number of cases we have added critical remarks, regarding the interpretation of the excavations and/or the radiocarbon dates.
In chapter 5 we analyse the 31 radiocarbon dates for decorated ‘Saxon’ pottery (all on cremated bone) and the 30 radiocarbon dates for undecorated handmade pottery of Hessens-Schortens type (partly on cremated bone, and partly on charcoal, unburned bone or wood), found in graves. By combining the limited amount of archaeological evidence and counting statistics we conclude that decorated ‘Saxon’ pots of Plettke’s type A6-A8 were used in burial contexts between c. 400 and c. 520, and that early ‘schalenförmige Gefässe’ of type C were used from c. 375 on. Archaeological evidence and radiocarbon dates indicate that decorated ‘Saxon’ pottery was used in settlement context during the same period.
Pottery of Hessens-Schortens type was used in graves between c. 500 and c. 730, but was already used in settlement contexts since c. 400.