Updated December 3, 2000
Model of dolmen
evolution
(3600-3290 B.C./
2860-2610 b.c. [3550‑3100 B.C.][8])
The estimated range of dolmen construction used for the model is 3600-3290 B.C. However, opinions vary. For instance, Bägerfeldt (1993:66, 68, 73) dates construction to ca. 3600-3350 B.C. and P. O. Nielsen suggests 3550-3100 B.C.
To understand chamber development during this time and later periods one must systematically analyze their dimensions, the evolution of the entrance, the number of side-stones, the number and position of capstones, as well as the orientation and location within long-mounds. This must be done in conjunction with analysis of the spatial distribution of the chambers throughout the entire culture area, not the arbitrary boundaries of individual countries or their modern subdivisions. The results lead to three classes of dolmen, each containing one or more chamber types (Table 2).[9] The chambers increase in size and architectural complexity through time, spreading from the core area into adjacent regions.
Primeval-dolmen (3600-3530 B.C.): True urdolmen are defined as having a megalith placed lengthwise on each side, a tightly fitting narrow stone between them at each end, sealed by a single capstone. Traditionally, these rectangular and trapezoidal chambers without an entrance are thought to be the oldest and it is commonly assumed that they were designed for a single articulated interment. This early custom supposedly differs from the later custom of multiple interments or collective burials derived from France along with passage-graves.
Perhaps a single interment was indeed the case at the relatively large Bogø By (closed?) urdolmen (Fig. 6), although some of the artifacts may suggest a post TRB burial.[10] But the closed urdolmen Hjortegårdene 9 contained mostly disarticulated human bones from a minimum of three individuals (Fig. 7).[11] This and other evidence may mean that even early closed dolmen were used for multiple interments, in which the earliest burials were pushed aside to make room for their successors[12] - a custom that occurs in later TRB chambers and gallery-graves, suggesting a long lived continuity.
Basic statistics show that all closed urdolmen in long-mounds, containing EN collared and lugged flasks listed by Knöll (1976), are oriented parallel to the long-mounds' axis.[13]
Transitional dolmen 3550-3460 B.C./2815-2750: The next stage is a rather innovative stage, initiated by the appearance of multiple side-stones and the use of a top-entrance at, let us say, 3550 B.C./2815 b.c. (There are no C14 dates.) The evidence for the development of top-entrances stems primarily from Germany (Fig. 9, 1-2), but it also occurs in Denmark (Fig. 8). The chambers still consist of two pairs of lengthwise placed side-stones, but one or two small additional stones provide room for the entrance plate. These chambers are still roughly parallel to the long-mound.
After the development of the top-entrance, chambers begin to show a tendency to be oriented perpendicular to the long-mound's axis.[14] For Mecklenburg one may stipulate an intermediate stage between top- and half-height front-entrance, as well as a possible phase without evidence of a passage (Fig. 9, 3-5). In other areas the location of open urdolmen within their mound and other evidence (Hoika 1990b) implies that some were still entered from the top of the mound by means of a shaft. But Schlicht 1979 suggested that the passages could also have been constructed of untraceable organic material, while the evidence of frequent EN long-mound remodeling (Liversage 1992, Rønne 1979, Zich 1995) means that the narrow primary mound and traces of a (non-megalithic?) passage may have been obliterated by the final mound layout.
The half-height front-entrance and evidence of passages become relatively common some time around 3530 B.C./2800 b.c. About the same time, square-dolmen with a similar entrance, confined to Sweden's west coast, must have been constructed, taking full advantage of locally available long stone slabs to maximize the available floor space.
Around 3500 B.C./2780 b.c. the nearly consistent use of 2+2 upright side-stones may have initiated the construction of extended-dolmen with single axial capstone. In areas where rectilinear forms remained the norm, builders took full advantage of the multiside-stone architecture by 3480 B.C./2760 B.C. They rotated the capstones from the axial to a perpendicular position and applied a yoke architecture, usually matching each of the two capstones with a pair of opposing side-stones. The obvious advantage of this architecture is that ever longer chambers could be created with minimal regard to the shape and size of the capstones. This simple, logical innovation led from extended- to grand-dolmen, so that grand-dolmen of rectilinear and polygonal form coexisted by ca. 3450 B.C./2740 b.c. as was the case at Rothensande 63, Schleswig-Holstein, where a rectangular mound contained both a six-sided polygonal- and a probable 3+3 side-stone grand-dolmen (Sprockhoff 1966).
The exact duration of extended-dolmen construction is difficult to ascertain (cf. Hoika 1990). Klokkehøj's two oldest collagen dates are 2610±90 and 2600±90 b.c. (K-2954a, b, Persson and Sjögren 1996 Appendix). Thus P. O. Nielsen (personal communication 1996) notes that "Klokkehøj has EN II pottery associated with the earliest burials being radiocarbon dated to 3340 BC. The funnelbeakers from Klokkehøj belong to Eva Koch's type V, which she dates to the EN II and MN AIa." Therefore, he would place the construction period into the "second part of the EN II, i.e. c. 34003300 BC." However, the model (Tables 1, 2) predicted that extended-dolmen with half-height entrance and two perpendicular capstones should have been built by 3480 B.C., which later turned out to match the date for Rastorf LA 1 (2760±45 b.c.).[15]
The model also suggest that grand-dolmen construction begins around 3450 B.C./2740 b.c., but P. O. Nielsen prefers a date of 3300 B.C. There are six dates for the extended/grand-dolmen Säve 57 (St-6957-6970, Persson and Sjögren 1996 Appendix). Of these only five apply, ranging from 2805±220 - 2590±95 b.c. They average to about 3405 B.C. As yet unpublished, Christian Lindqvist and Göran Possnert performed new human collagen C-14 dates from the extended/grand-dolmen of Tofta, on Gotland (Bägerfeldt 1992). They date to about 3400 - 3300 B.C. (Persson Internet communication July 1996). One would assume, that the custom of grand-dolmen construction would have reached the Swedish west coast and especially far off Gotland after its inception farther south, so that these dates could also fit the model. At any rate, Ebbesen (1979:71, 102-103) observed MN A Ia as earliest pottery in some grand-dolmen and MN AIb in others, suggesting the coexistence of these chambers with passage-graves. In deed, the largest three- to four-yoke dolmen were probably built well into the period of extended-passage-grave construction, which, as the model suggests, began around 3290 B.C./2610.
The innovation of upright side-stones permits easy adjustment to the shape of the best available capstone, resulting in extended-dolmen that sometimes took on a polygonal form around 3480 B.C. This process is seen in chambers such as Hyllinge, the extended-dolmen-like chamber with perpendicular capstone of Bregninge, and the polygonal-dolmen of Nødager (A. P. Madsen 1868:16 No. 25, 1900:11 No. 4, 6). The polygonal shape became favored in parts of the North Group.[16] The continued demand for increased chamber size thus also led to polygonal grand-dolmen by perhaps 3450 B.C./ 2740 b.c.[17]
However, in the literature the classification of polygonal chambers is imprecise. Some are simply odd shaped extended- and grand-dolmen. Others are probably early oval passage-graves. In fact, Lyse 7 on Sweden's west coast is identified as a passage-grave due to its long passage (Bägerfeldt and Kihlstedt 1985), even though its shape and size suggest an early polygonal chamber. To complicate matters, it's sole C14 date (St-4567: 3595±155 b.c., Persson and Sjögren 1996 Appendix) is Mesolithic. However, there are two consistent dates from Tustrup, Jylland, of ca. 3466 and 3441 B.C. (KL-1762: 2750±110, KL-1763: 2730±110 b.c., Persson and Sjögren 1996). The poorly preserved polygonal (?) chamber Holtenes III is dated 3414 B.C. (T-5828: 2710±45 b.c.), which seems appropriate, since the model expects the construction of tombs in Oslo Bay to be slightly later. Yet caution is advised, for P. O. Nielsen (personal communication 1996) sees the destroyed Danish chamber Trekroner as a possible polygonal-dolmen. It is dated to the early MN AI (Lu-1952: 2550±55 b.c., Persson and Sjögren 1996 Appendix). Noting the occurrence of wiggles in the C14 curve, he suggest a duration of 3400-3250 B.C. for polygonal-dolmen construction.
Related Links
Neolithic/Copper
Age Link Index: Links to News Bulletins, Articles, Site Reports,
Databases, etc. about the Neolithic/Copper Age in Europe.
Please send comments or questions to Max Baldia.