by Chris Fowler
Middle Neolithic
In Ireland ceramics are marked by a series of regional sub-styles and modified forms developing after 3600 BC (Sheridan 1995, 18). In Britain the important transition is the development of the broad Impressed Ware tradition, incorporating the Ebbsfleet, Mortlake, and Fengate sub-styles of Peterborough Ware in the south and east with other regional groupings further north. Since the review of radiocarbon dates by Gibson and Kinnes (1997) it is clear that the Peterborough styles at least begin around 3400 BC and that there is no chronological significance to the sub-styles. Similar conclusions were reached for Welsh Peterborough Ware by Gibson (1995, 30), with all pottery dating 3500–2500 BC. In central and eastern Scotland Impressed Ware dates tend to cluster in the earlier part of the third millennium (Cowie 1993, 21). In the west of Scotland Sheridan (2000) has suggested dates between 3500 and 3000 BC for Hebridean Wares, with Orcadian Unstan Bowls and the Collared Bowls of the south-west belonging between 3500 and 3250 BC.
Late Neolithic
The late Neolithic is dated by the introduction of Grooved Ware. This did not take place at the same time in all parts of Britain and Ireland, and the relative importance of Grooved Ware assemblages varies from region to region. Key work on the dating of British Grooved Ware by Brindley (1999b) and Garwood (1999) puts the style as a whole between 2900 and 2100 BC. It is also now clear that Grooved Ware has its origins in northern Scotland (Garwood 1999, 146), probably deriving from the middle Neolithic Unstan Ware of the Orkney Islands. Grooved Ware in Ireland has only been identified relatively recently (Brindley 1999a); here it is a relatively rare component alongside the continued use of modified bowl pottery and was probably introduced around 2900–2700 BC.
Summary
Over the whole of Ireland and Britain it would seem sensible to discuss pottery developments using the following four-phase framework: first Neolithic 4000–3800 BC; early Neolithic 3800–3500 BC; middle Neolithic 3500–2900 BC; and late Neolithic 2900–2400 BC. Although these phases are averages of regionally specific events and changes they should give a dated structure for major changes in the way people made and used pottery. The following two sections discuss the surviving evidence for the manufacture and use of pottery using this chronology (see Fig. 30.8).
PRACTICE AND MATERIALITY
Pottery, like all artefacts, arises out of the application of knowledge and decisions. We recognize and reify the results of these decisions as our named styles of pottery but it is likely that categories of practice and material engagement in the prehistoric past were both more fluid and less conscious than this. The following section reviews the evidence looking primarily at how people worked with clay and used pottery.
Unfortunately, clay is such a plastic material that unambiguous traces of specific practices can often be obscured. This has particular relevance for debates about the transmission of pottery-making knowledge. Attempts have been made in lithic studies to identify assemblages characteristic of the process of learning knapping (Finlay 1997; Pigeot 1990). Nothing similar can be attempted with ceramic assemblages. If pottery seems to us ‘crude’, ‘experimental’, or ‘unskilled’ we must remember that it was nonetheless chosen for firing and use in the Neolithic: the place for unsuccessful experiments would have been the reclaim bin—or at least its prehistoric equivalent—rather than any context which might preserve them for our study. Similarly, we should not be surprised at a high degree of technical competence in the earliest surviving pottery.
There is extensive ethnographic literature describing how pottery has been made and used in traditional societies around the world (e.g. Van der Leeuw and Pritchard 1984; Rice 1987). Taken in conjunction with detailed technical analysis of Neolithic ceramics (e.g. Darvill 1984; Jones 1999), it is possible to use this to suggest traditions of manufacture and use and to try and ascribe dates to them.
Clay and inclusions
A source of workable clay is a fundamental requirement for all potters. Ethnographic surveys (Arnold 1981, 35) suggest that most modern potters seek clay within a single day’s travel of the production site, although Rice (1987, 117) points to a much wider variety of practices. Depending on the geochemical properties of the source the clay may then need to be modified by processes such as levigation (mixing clay with water to allow impurities to settle out), souring (storing wet clay to allow water permeation and bacterial action to increase the plasticity), wedging (kneading the clay to homogenize it and remove air pockets), and mixing with other clays. Many modern potters add inert material (temper) to their clay (see Rice 1987, 118–24, for a survey of varying examples). This tends to have two functions: it aids the firing of the clay, promoting a variety of chemical reactions during the firing process (see Rice 1987, 93–8 for a summary) and greatly increases the chance of a pot surviving the firing; and, depending on their properties and size, inert inclusions can modify the physical properties of the finished vessel, making it either more robust or better able to withstand the stresses of heating and cooling (see Steponaitis 1984, 85–108, for a detailed technical discussion of these effects). All of these processes can potentially leave traces in the body of the finished pot which can be detected, particularly through the use of thin section slides (see Peacock 1970 for a review of methodologies). Historically, thin section data has tended to be used as a tool for sourcing pottery production (see Riley 1984 for a review). Studies also exist which allow us to see the way clays and inclusions from different areas were chosen and blended in the same assemblage; a good example being analysis on the pottery from Trefignath chambered cairn on Anglesey (Williams and Jenkins 2005).
Manufacture and firing
Ethnographic examples of known techniques for hand-building pottery are surveyed by Rice (1987, 124–8). Very detailed formal accounts of specific hand-building traditions are presented by Krause (1984) and Van der Leeuw (1984). It is very difficult to be sure exactly how people built pottery in the Neolithic: an expert hand potter can, if they desire, almost completely remove all traces of the processes used to form a vessel. Fortunately, some data do survive within vessel walls from most British Neolithic assemblages. This usually takes the form of voids left by the incomplete joining of one segment of the pot to another (see Fig. 30.9). Woods (1989, 196–200), reviewing this evidence, suggests that the dominant way of working was to build vessels from cylindrical straps of clay rather than the annular rings favoured by modern craft potters.
FIG. 30.9. Cross-sectional views of Neolithic pottery from Blasthill chambered cairn, Argyll, showing voids and breaks in the fabric along the joins between segments.
For pottery to be any use it must be fired. No well-defined Neolithic pottery production sites have yet been found, and very few assemblages have traces of production evidence—the kneaded but unshaped clay noted from Biggar Common West, South Lanarkshire (Sheridan 1997, 202) is one example. Generally, only the data from the pottery itself tell us about firing practices. It is clear from the absence of even simple kiln-type structures that all firings were a variant on the simple bonfire and a wide range of experimental work has been carried out illustrating its effectiveness (e.g. Woods 1989). Clays pass through two chemical changes in the temperature range we might expect for open fires: dehydration and oxidization. Dehydration of the clay minerals occurs as they are converted from hydrous and hydrated forms of aluminium silicate to anhydrous forms. The clay looses its plasticity as the crystalline structure of the clay minerals breaks down and particles are no longer free to move against one another. Iron and carbon compounds are also present in clays as impurities. Oxidization of the carbon compounds takes place at lower temperatures than of the iron compounds and it is the incomplete removal of this carbon which give the characteristic dark core to partially oxidized sherds (Shepard 1968, 20–21). Carbon oxidization depends on fabric porosity, kiln atmosphere, firing temperature, firing duration, and the rate of temperature increase. The last of these can be disregarded here since open bonfires produce similarly rapid rates of
temperature increase (Shepard 1968, 74–91). Fabric porosity is difficult to measure except by techniques which would be destructive to prehistoric ceramics (Shepard 1968, 127). The temperature reached during firing can be partly inferred from physical changes in clays and some inclusions, particularly if thin-section data are available (Peacock 1977, 30; Steponaitis 1984, 83; Tite 1972, 230). Kiln atmosphere and firing duration both depend on the type and amount of fuel used (Hodges 1962, 63; Rye 1981, 102; Shepard 1968, 79–84 document extensive experimental work on the effects of these choices), but also on whether or not a pit is used for the firing. Therefore, using the data available from assemblages studied in detail, it ought to be possible to make broad inferences about different firing techniques; particularly where thin section analysis has been carried out (for example the analysis by Smith and Darvill (1990, 151) of pottery from Hazleton North).
Use
The past uses of archaeological ceramics can be analysed through a range of well-established techniques which work best where vessels have consistently been used for similar tasks. At a very simple level is the recording of fire traces on the exterior of vessels used for cooking, creating a paler and often pitted surface on the base of a pot used in this way (Fig. 30.10). It is possible to identify vessels with this secondary oxidization and only slight sooting on the upper walls; these would have been habitually used for cooking by placing the base directly on a bed of embers. Other vessels display a different pattern of intense sooting of the base and little or no secondary oxidization; this implies these vessels were usually suspended above the fire during cooking (Rice 1987, 235). Remarkably few Irish and British Neolithic vessels show no traces of cooking at all.
FIG. 30.10. Base and wall sherds of a complete Neolithic vessel from Blasthill, Argyll, showing secondary oxidization of the base and sooting on the walls. This vessel was placed directly on a bed of embers as a cooking pot.
Charred food remains survive in a surprisingly high proportion of Neolithic pottery but the most successful analytical technique for studying the use of pottery has been gas chromatography of food lipids which have become chemically bonded with the clay (see Evershed et al. 1992 for a review of methodologies). This has been used both to demonstrate consistent links between different vessel types and different foodstuffs (Jones 1999), and to document changing diets over a longer time period (Dudd and Evershed 1999). It is also possible to make broad statements about potential vessel use from the known physical characteristics of the fabric (Steponaitis 1984, 95–116). Of particular relevance are the mechanical strength of sherds and the ability of the pot to withstand thermal shock and stress. A study of this kind by the author established that these categories did not appear to limit the use of Neolithic pottery for cooking, although they may influence the kind of cooking that is possible (Peterson 2003, 151).
MEANING AND ASSOCIATIONS
The first Neolithic pottery
During the first part of the Neolithic it could be suggested that pottery production was strongly bound by rules and procedures. This can be most clearly seen in the relatively restricted range of vessel shapes and the absence of decoration in most assemblages (see Fig. 30.11). By contrast, where we have thin-section data from dated assemblages, there is a tendency for clay and inclusion ‘recipes’ to be complex—as at Trefignath, Anglesey, for example (Smith and Lynch 1987; Peterson 2003, 52–4). This complexity seems to be both within particular assemblages and across regions. Clay and inclusion choices are much more complicated than the minimum technical requirements to successfully fire and use pottery.
Despite the ‘fine’ character of much of this early pottery it appears to have been used for cooking. Although Herne (1988, 26) describes it as having a special character and role on the basis of his survey of British and Irish material, he also concedes that this was probably bound up with its primary function as a new technology of food preparation. Certainly large assemblages with early dates, such as from Biggar Common West, South Lanarkshire (Sheridan 1997, 204), show clear evidence of use for cooking on open fires.
FIG. 30.11. Selected dated material belonging to the first Neolithic: (a) Ballymacaldrack, Antrim (after Herity 1982); (b) Biggar Common West (after Sheridan 1997); (c) Gwernvale (after Britnell and Savory 1984); (d) Flagstones (after Cleal 1997); (e) Tankardstown South (after Sheridan 1995).
Developing variability
By the later part of the early Neolithic these clay and inclusion recipes seem to have become slightly more standardized. However, unusual tempering materials such as the burnt bone found in vessel 32 from Hazleton North, Gloucestershire (Smith and Darvill 1990, 152) also seem to belong to this period. Forms remained relatively standardized but in particular regions decorated assemblages developed (Fig. 30.12). Although Herne’s (1988) model would see variability developing from a standardized ‘carinated bowl’ at about this date in some areas (Wales for example; see Peterson 2003, 152) the range of vessel forms actually narrows. Analysis of the potential and actual uses of early Neolithic pottery from Hazleton North by Smith and Darvill (1990, 151) seems to show distinct classes of cooking and storage jars in the assemblage.
FIG. 30.12. Selected dated material belonging to the early Neolithic: (a) Gwernvale (after Britnell and Savory 1984); (b) Hazleton North (after Smith and Darvill 1990); (c) Pitnacree (after Cowie 1993); (d) Windmill Hill (after Zienkiewicz and Hamilton 1999).
Impressed Wares
By the middle Neolithic it seems as if traditions of pottery production had undergone a major change. Now the emphasis was on a wide variability in vessel form, especially in the areas outside the distribution of the Peterborough sub-styles (Fig. 30.13), allied with a profusion of new decorative techniques common to all the Impressed Ware traditions. The other major change at this date was in the inclusions used. Impressed Ware assemblages, despite the variability in form and decoration they display, tend to be tempered with locally available materials. This temper is usually processed in a very perfunctory way, leading to substantial chunks of rock inclusions such as quartz being present in much pottery of this date; see for example the large Mortlake Ware assemblage from Ogmore-by-Sea, Glamorgan (Gibson 2001).
A much more detailed understanding of the use of middle Neolithic pottery was possible from organic residue analysis carried out at Upper Ninepence (Dudd and Evershed 1999). These vessels had clearly been used for cooking meat, primarily from cattle and sheep. Similar analysis on an Unstan ware assemblage from Stonehall, Orkney (Jones 1999, 63) demonstrated that the pottery had been used for cooking both cattle meat and milk.
FIG. 30.13. Selected dated material belonging to the middle Neolithic: (a) Goodland (after Sheridan 1995); (b) Northton, Harris (after Murphy and Simpson 2003, fig. 11,4); (c) Ogmore-by-Sea, Glamorgan (after Gibson 2001, fig. 6).
Grooved Ware
By the late Neolithic this utilitarian approach to pottery temper seems to have been further simplified by the widespread introduction of grog tempering. We must be careful not to overstate the absence of grog in earlier periods as where we have thin-section data grog appears to be a part of even the very complex tempering strategies of the first and early Neolithics (Darvill 2004, 194). However, by the late Neolithic we see assemblages primarily tempered with grog or with single stone types, such as the assemblage from Links of Noltland, Orkney (Sheridan 1999, 116). Whilst vessel form and decoration shows a similarly wide range of possibilities as in the middle Neolithic, there is much more sense of these choices being widely disseminated and understood over large areas. This is reflected in the established Grooved Ware sub-styles which cover the whole of the study area (Fig. 30.14). Organic residue studies have established a range of potential uses for Grooved Ware pottery. From Barnhouse, Orkney, there is evidence for the cooking of beef, milk and barley (Jones 1999, 64), whereas at Upper Ninepence the pottery seems to have been for cooking pork (Dudd and Evershed 1999, 117). Charred remains and pollen within Grooved Ware at Balfarg indicate cereals, potherbs, and some more exotic and
potentially potent plants such as nightshades and Solanaceae (Moffat in Barclay and Russell-White 1993, 109).
FIG. 30.14. Grooved Ware and other late Neolithic pottery: (a) Lambay Island (after Sheridan 1995); (b) Links of Noltland (after Sheridan 1999); (c) Upper Ninepence (after Gibson 1999); (d) Balfarg Riding School (after Barclay and Russell-White 1993).
CONCLUSIONS
In summary we can see Neolithic pottery in Ireland and Britain as a technology made up of a number of traditions and practices. The shape of vessels seems to be strongly codified in the first instance, although the range of different inclusion recipes would suggest an experimental radiation of practices. In the middle and later Neolithic we can suggest there was a simplification of the technical aspects of potting alongside a radiation in the range of shapes and uses that were thought appropriate.
