The Orkney Grooved Ware Experiment

During April 2007 a small group of volunteer potters, led by archaeologist Stephen Harrison and professional potter Andrew Appleby, began a three-year experimental research project into the making of Neolithic pottery. The project, an active collaboration between archaeologists and potters, with substantial community involvement, is based at Fursbreck Pottery, Harray, Orkney, and aims to come to some potential understanding of the processes involved in the manufacture and firing of prehistoric ceramics and their use during the third and early second millennia BC within a specifically Orcadian context.

Click below to download the Abstract

Orkney Grooved Ware experiment

If you would like to contact Stephen his email address can be found in the abstract.



© Text and images copyright Stephen Harrison

Plate 1. Digging clay at Stackle Brae on the south coast of Eday, September 2006.

Plate 2. Close-up of clay from Clay Loan, Kirkwall, showing abundant sedimentary inclusions. In terms of Orkney clays, this material is very suitable for potting.

Plate 3. Initial sorting of clay.

Plate 4. Soaking clay. The wheelbarrow contains the separated larger naturally occurring inclusions.

Plate 5. Discarded material from the initial sorting of clay. When dry, this material forms a compacted surface. Can such features be recognised in the archaeological record?

Plate 6. Kneading clay prior to potting.

Plate 7. Kneading shell sand, as a tempering agent, into clay prior to potting.

Plate 8. Organic material – chopped damp grass – as a tempering agent. The dampness of the grass greatly enhances the plasticity of the clay

Plate 9. Making a small cup (‘pinch pot’) from a 250gm ball of clay. Depending on the degree of surface finish, these vessels can be made very quickly: between five and twenty minutes each.

Plate 10. Using a 2-3kg ball of clay, small and medium-sized tubs can be made in a single operation within thirty minutes. The makeshift turntable is an ideal aid to manufacture, allowing the evolving vessel to be rotated at will.

Plate 11. A 10kg block of clay ready for potting.

Plate 12. Using a smooth, rounded beach pebble to hammer out base and lower walls.

Plate 13. Drawing up vessel walls using a pinch-and-pull technique. Drawing the vessel walls up in this manner tends to orientate inclusions vertically on their long axes. On the other hand, inclusions in coil-built pots are randomly orientated

Plate 14. Base and lower wall completed. From the initial block of clay, this task was achieved in twenty-six minutes.

Plate 15. Adding coil. As the vessel grows in height, the positioning of coils is important in determining the ultimate profile: for vertical-sided vessels (as above), coils need to be placed more on the interior side of the preceding section; conversely, if a angled profile (flower pot shape) is required, coils should be attached more on the exterior side.

Plate 16. Pressing coil down to attach to already formed wall.

Plate 17. Having attached the coil to the wall, it is then drawn up. Here, a false rim is being created to attach the next coil. The area where the coil was joined to the base section can clearly be seen at the bottom of the photograph.

Plate 18. Using a pebble to smooth vessel surfaces.

Plate 19. Once the vessel has reached the required height, a rim coil can then be added.

Plate 20. Creating a scalloped rim using a section of sheep rib bone as a makeshift tool.

Plate 21. Using thick clay slurry to plaster vessel surfaces. When dry this provides additional structural support as well as masking any imperfections from the construction process.

Plate 22. When the plastered surface is dry it can be smoothed using an obliquely cut sheep rib bone.

Plate 23. Decorative scheme roughly marked out. This was achieved using a bone pin.

Plate 24. Applied decoration. Very fine, inclusion-free clay is needed for producing applied decoration

Plate 25. The finished vessel. This pot measures H. 30.5cm x Mouth W. 31.8cm x Base W. 26cm x Wall Th. 1.2cm. Built from 15kg of clay tempered with chopped damp grass, the vessel took 456 minutes to complete: 23 minutes kneading clay and adding tempering agent + 125 minutes building + 207 minutes resting time during the building stage + 101 minutes decorating.

Plate 26. Applying incised decoration using a bone pin. This vessel took fifteen minutes to decorate.

Plate 27. Durrington Walls-style all-over incised decoration. This tub (H. 18cm x W. 19cm) was made from a 5kg block of clay with chopped grass as a tempering agent. The vessel took thirty-nine minutes to build, but 118 minutes to decorate. Interestingly, after two abortive attempts at decorating the vessel in an upright position, it was found that the decorative scheme could only be successfully achieved by inverting the pot and working from the base downwards.

Plate 28. Vessels air-drying immediately after manufacture.

Plate 29. The potter’s tool-kit: an aspect of material culture invisible in the archaeological record.

Plate 30. Building the first turf kiln.

Plate 31. The turf kiln ready for use.

Plate 32. The first stage of the firing process. Once the flames from the wood fire died back, the glowing embers were spread across the base of the firing chamber and covered with a layer of peat mould. After about ten minutes, the peat began to smoulder; loading vessels could then begin.

Plate 33. Loading the first layer of vessels.

Plate 34. Loading the kiln. Owing to poor quality clay (very high sand content in the raw material), the large vessel (centre), H. 59.7cm x W. 53cm, did not survive the firing.

Plate 35. Kiln loaded, capped with a domed layer of wet seaweed, and firing. A slow rise in temperature means that vessels are better able to withstand the adverse effects of thermal shock.

Plate 36. The first vessels appearing in the spent fuel ash. Although not visible on this photograph, the large vessel has numerous deep cracks across the base and along the walls, and fell apart during removal from the kiln.

Plate 37. Unloading the kiln: a handful of straw provides good insulation for this task.

Plate 38. The kiln partially unloaded.

Plate 39. A range of fired Grooved Ware vessels.

Plate 40. Montage of the area immediately outside the kiln after unloading (kiln at top of photograph), comprising a trampled and compacted surface of broken and crushed pot sherds, fuel ash, and blocks of turf superstructure.

Plate 41. Close-up of heat damaged kiln wall. This material is very similar to the bulk samples obtained from the Knowes of Trotty structure in 2006.

Plate 42. Excavating the kiln, August 2007. The multi-coloured ash deposits were between c.0.3-0.4m in thickness, beneath which was an uneven basal layer of charred wood and charcoal from the initial pre-loading fire.

Plate 43. The base of the firing chamber after excavation. On sectioning, it was found that heat penetration into the underlying surface was minimal (2-5cm).

Plate 44. Magnetometer survey of the kiln’s footprint by members of the Geophysics Unit, Orkney College/UHI, August 2007.

© Text and images copyright Stephen Harrison





Site designed and maintained by