PETROGRAPHIC ANALYSIS OF NEOLITHIC POTTERY ON ORKNEY - NEW FINDINGS

Lara Maritan* and Richard E. Jones**

*Department of Mineralogy and Petrology, University of Padova, Italy
**Department of Archaeology, University of Glasgow

The Neolithic pottery of Orkney is well known: round-based pottery (Unstan ware) is typically associated with the Early Neolithic and flat-based pottery (Grooved ware) with the Late Neolithic, and yet debate continues as to their significance beyond their traditional role as indicators of a cultural sequence. Owing to their coarse fabrics, petrographic analysis is ideally suited to the investigation of their material aspects, as Andrew Jones has recently shown most effectively in his study of the dynamics of Grooved ware production at the settlement at Barnhouse on Orkney (Jones 1997; 2000). Both his work, and that of David Williams (1979; 1982) carried out over twenty years ago, demonstrated the potential importance of the presence of igneous rock fragments in the early pottery from Orkney.

Excavation of two newly discovered, neighbouring Neolithic settlements on Mainland Orkney, at Stonehall and Crossiecrown, situated near Cuween tomb and Quanterness respectively (Downes & Richards 2000), has stimulated a new programme of petrographic study carried out as a cooperative effort with Andrew Jones (Southampton University) and Colin Richards (Manchester University); the preliminary results have been reported by Maritan et al. (2000). At Stonehall the crucial point of interest was the long occupation of the settlement, probably throughout the Neolithic, and the occurrence therefore of both classes of pottery at the site. On the basis of 92 representative examples of Round-based ware and Grooved ware from the houses excavated at this site, one of the main questions that the petrographic analysis tackled was the similarity or otherwise of the materials used in pottery production over the course of the Neolithic.

Fig 1

Fig 2

On the basis of the nature of the large inclusions, two main groups have been identified. The largest is characterised by the presence of millimetric angular to well-rounded fragments of sandstone, siltstone, argillite and silty crystals of quartz, plates of white mica, biotite, plagioclase, K-feldspar and opaque minerals (Fig. 1). In addition to these sedimentary rocks and minerals, the other group also contains a more or less abundant quantity of magmatic fragments of camptonite or dolerite and strongly weathered rock of magmatic origin (Fig. 2). While the sedimentary inclusions are not diagnostic of origin since the Orkney Islands are mostly composed from Old Red Sandstone (Mykura 1976; BGS 1999), the magmatic fragments by contrast allow one to isolate their possible source areas. In particular, with the aid of the British Geological Survey's thin section collection of magmatic dykes and bodies on Orkney, it was possible to identify that camptonite was locally available. The same may apply to dolerite even though no dykes have been found in the vicinity of Stonehall. Instead it is likely there was a local outcrop of dolerite because the results obtained from Crossiecrown indicate that in a few examples of pottery the presence of this rock is associated with others of surely local origin. As regards the presence of these large inclusions in the pottery, petrographic comparison with clay material collected on Mainland Orkney (Jones & Brown 2000) suggests that the large sedimentary inclusions were natural component of the boulder clay, the predominant clay material available on Orkney. The magmatic fragments, which often do not exhibit angular shapes, could represent either the natural component of a boulder clay collected close to the dykes, or a temper material coming from the disaggregation of the magmatic outcrops. Thus the question whether these sherds were tempered or not is still open.

The two main petrographic groups at Stonehall have no apparent chronological significance, the same range of materials being represented in both Round-based and Grooved wares. On the other hand, there is evidence of a spatial dimension within the petrographic classification; that is, pottery from the different houses have somewhat different compositions in a manner that suggests that, as at Late Neolithic Barnhouse, the potters in individual houses were perhaps using their own recipes. These finding certainly have important implications for the interpretation of the settlement, as well as the organisation of ceramic production there.

Fig 3

At Crossiecrown, from which 34 samples were examined, the microscopic analysis indicates some contrasts with nearby Stonehall. While some of the pottery contains only sedimentary fragments, such as sandstone, siltstone and argillite, most of the sherds examined also contain inclusions of magmatic rocks; not only is there camptonite, but examples of dolerite, monchiquite and basalt as well. Intriguingly, there are a few instances of the concurrence of these magmatic rock types (all of which were probably locally available) within individual sherds (Fig.3). This should constitute strong evidence for delibrate tempering, whereas other pottery from the site seems to have been produced using boulder clay without the addition of temper. As at Stonehall, the indications hint at individual potters exerting personal choice in the selection of the raw materials they used.

References

Back to meeting programme

Back to main menu