Swifterbant, Oost Flevoland, Netherlands: Excavations at the river dune sites, S21-S24, 1976. Final reports on Swifterbant III
Continuing investigations in the Swifterbant area over the last twenty years have resulted in the recovery of the remains of human occupation dating from both the Mesolithic and the Neolithic periods. The somewhat higher elevations of the river dunes appear to have been favored locales for habitation. During the Boreal period of the early Postglacial, this area would have been a fairly typical coversand environment in a riverine regime. The landscape would have been dominated by a relatively stable Boreal forest. As conditions became moister and sea level rose after 5500 B.C. in the Atlantic period, the area was transformed into a bog and creek system in a fresh water tidal delta environment. A marsh forest would have been present with much alder, oak, and pine. Lime trees would have been common on the river dunes (Casparie et al., 1977). Finally the area would have been inundated by rising water levels sometime after 3000 B.C.
Although much of the occupation horizon on the river dune at H46 has been truncated by erosion, some of the materials collected from the site appear to document a Ceramic Mesolith ic occupation. Ceramic Mesolithic sites are also recorded along the Atlantic and Baltic coasts of Europe. In northern Germany and southern Scandinavia pottery is found with the Ertebølle-Ellerbek cultures of the Mesolithic. Although organic materials are not preserved and macrolithic tools such as flint axes are missing in the river dune sites at Swifterbant, the date of 4300 B.C. at S23 falls within the range of the Ertebølle of Denmark which begins around 4600 B.C. (Brinch Petersen, 1973). Ceramics in the Danish Ertebølle, similar in size and shape to the Swifterbant material (de Roever, 1979), appear after 3800 B.C. in southern Scandinavia. In France, ceramics in association with Mesolithic remains have been reported along the Atlantic coast at Roucadour (Roussot-Laroque, 1977). This material is dated to approximately 4000 B.C. and the ceramics are similar to the Swifterbant material with pointed bases and heavy walled construction. The evidence from these areas supports the earlier hypothesis of Schwabedissen (1966) that ceramics may have diffused into northern Europe along the coast prior to or simultaneously with the appearance of inland pottery-using farming cultures.
An earlier utilization of the dune is indicated as well. Radiocarbon dates of roughly 5800 B.C. and 4800 B.C., along with certain lithic artifacts, provide evidence for earlier Mesolithic occupations, corresponding to the Boreal and Late Mesolithic periods of Newell (1973). The graves on the dune, intrusive through the developed soil horizon, must postdate 4300 B.C. These graves are likely from the end of the period of potential occupation on the dune. The facts that the graves are intrusive through the soil and that the bones of the burials are partially preserved argue for interment at a time when water levels were higher in the area. These graves may well be contemporaneous with the Neolithic settlements and burials from the clay levee sites in the Swifterbant area. The levee burials are dated to 359 B.C. by a radiocarbon date on bone collagen from a grave in parcel G42 (GrN -5606 5540±65 B.P. ; Van der Waals & Waterbolk, 1976, p. 7).
The graves at H46 show a distinct orientation (figs. 4 & 7). With only one or two exceptions, the longitudinal axes of the graves follow the top of the dune and are located near the former crest.
From the angle of the north and south slopes of the dune (fig. 3), the crest of the dune must have stood at least 75 cm higher than its present surface, or at approximately - 3.75 m N.A.P. The orientation and placement of the graves thus also argues for the original location of the graves in the highest and driest areas available and at a late date in the period of potential occupation.
Although one of the major goals of the 1976 project, the location and recovery of an intact Mesolithic occupation surface was not achieved, the season was successful. Earlier excavations at the dune were completed and an additional burial removed before it was destroyed by the drying of the dune. The new excavation unit, S23, completely transected the dune and provided valuable information on the occupational history of the site.
In essence we have learned that the higher areas of the dune have been badly disturbed. In areas along the slopes of the dunes, two probable cultural horizons are present and vertically separate, albeit with some overlap. Artifactual material, vertical separation, and radiocarbon dates argue for a possible Ceramic Mesolithic in the Swifterbant area. However, because of the disturbance of the dune and the overlap of the cultural horizons, successful recovery of an intact surface from this period may be possible only in deeper excavations - where the water table will make discovery difficult - or on those lower dunes (e.g., S11) that were not truncated by marine erosion. Clearly, more investigation is needed with better stratigraphic separation to adequately document the settlements of the pre-Neolithic inhabitants of the Swifterbant area.