VARIABILITY.

Since the mid-1960's, when New Perspectives in Archaeology and Analytical Archaeology were published (Binford and Binford, 1968; Clarke 1968), numerous artefact studies have been published examining factors other than date and 'culture' which may affect the composition of archaeological assemblages. Few of these studies have dealt with the Saxon or medieval periods in the British Isles but there is no reason why such studies should not be relevant here. The possible factors suggested for other cultures should be examined to see whether they may affect the results of this thesis.

It has been proposed that the variety of artefact types available in a society is related to the social differentiation, since artefacts are used to express social differences. This undoubtedly applies to the medieval period. It is known that dress, for example, was used as an indicator of status and there were at times sumptuary laws to restrict the wearing of particular types of dress except by people of the appropiate rank.

It is possible that cooking and storage vessels would be affected by this attitude but it is more likely to have influenced the choice of pottery types used for serving, drinking and display. Of these, the obvious types to examine are the decorated jugs of the 12th to 14th centuries and the decorated tableware of the late 13th to 17th centuries. The earliest highly decorated jugs found are of late 12th century date (for example, some tubular-spouted pitchers) but these are so rare within the region that it is not possible to make deductions from their occurrence. In the 13th century highly decorated wares are found, for example Ham Green ware and Worcester-type jugs. No relationship between their occurrence and the type of site has been noted, nor any significant variation in the relative frequency of the plainer forms to the 'highly decorated' vessels. Highly decorated Ham Green jugs are found on deserted medieval village sites, although they might have been used in the Manor House rather than the peasant tofts.

In the later 13th and 14th centuries the quantity of highly decorated jugs in use was at its peak and included some of the highest quality pottery of the medieval period. One such type was Saintonge ware, and especially the polychrome decorated jugs. It has often been stated that the distribution of Saintonge polychrome ware is strongly correlated with that of seigneurial sites, such as abbeys and castles. The distribution within the region is examined in Chapter 11 but it seems that here the most significant factor is distance from the coast (not distance from navigable waterways). This suggests that the vessels were transported on sea-going ships which could not travel up rivers like the Severn and that Saintonge polychrome ware was not considered valuable enough to be transshipped. The apparent seigneurial distribution is, in the writer's opinion, merely a reflection of the concentration on castle and abbey excavation in the region.

The distribution of other fine tablewares in the medieval period is more promising. Andalusian lustreware is absent from the region but a single bowl is known from Devizes Castle. Valencian lustreware is uncommon and the few sherds from Gloucester and Worcester (Morris, 1980) might actually date from the 16th century, by which time the quality of the ware, and its contemporary value were probably lower than in the medieval period. A set of 15th-century Valencian lustreware vessels was found at the Pithay in Bristol and this surely must have been the property of a wealthy or high status person (Hurst,1977a, nos. 18-19, 23-24, 27). Later decorated tableware can be divided into five groups;

Imported continental earthenware (for example Beauvais sgraffito, North Italian, Weser and Werra slipwares)
English copies of these wares (for example South Somerset slipware)
Chinese porcelain
Italian maiolica
Dutch and English copies of these wares.

Before the mid-17th century groups 1, 3 and 4 are extremely rare in the region but the few examples of group 1 come predominantly from sites close to the coast, navigable rivers or from sites in towns with known trading connections. A single example of Italian maiolica is known, from Hailes Abbey (and associated with the house built on the site of the Abbey at the dissolution). No 17th century Chinese porcelain is known from the region and late 16th to 17th century Dutch or English tin-glazed ware is also rare (examples are known from Cirencester Abbey, another site where a large country house was built after the dissolution).

The conclusion of this survey is that pottery was probably used to express social differences in the post- medieval period and to a lesser extent in the medieval period or to be more exact there is a greater range of 'finewares' in the post-medieval period. However, most of the sites examined have a very similar range of 'fine wares'. These sites range in status from castles and abbeys to deserted medieval villages. The relative frequency of jugs to other vessels differs between sites but the absence of stratification on most rural sites makes this difficult to quantify. However, it is likely that if this does represent a social difference it is not in the ability to purchase glazed jugs but in the use by the richer members of society of metal cooking vessels, so giving rise to lower quantities of cooking pot sherds on their occupation sites.

Pottery was used for a variety of purposes (see chapter 7) and one might expect assemblages to vary in the relative proportion of vessels of different function. However, most assemblages contain a mixture of functional types many of which would probably not have been used together. Therefore the relationship between area of use and type of vessel has been lost. This must be due to the methods of rubbish disposal. It has recently been claimed that spatial differentiation exists in the medieval pottery from a site in Southampton (M. Brisbane, pers. comm.). In this case the ratio of cooking pots to others increased with distance from the living quarters. Some well-groups have been found in the region which likewise show functional grouping - presumably these vessels were lost whilst being filled with water (for example, a mid-13th century well group from site 53/69 in Gloucester is composed almost entirely of jugs, as is the Bristol Castle well group, Barton 1959). The excavation of the hall and chapel of Holm Castle, Tewkesbury, produced assemblages dominated by glazed jugs with very few cooking pots. Although this might be due to the wealth of the inhabitants, who might have been able to use metal instead of ceramic cooking pots, it might also be that the kitchen area of the castle was not excavated and if found would have redressed the balance and proved that intra-site variation in the distribution of pottery exists.

These examples are rare and emphasise how similar the remaining assemblages are. There are differences in function between groups but these are mainly explicable as country-wide changes through time and do not lead to any greater understanding of individual sites. If this is so, then it is likely that they do not severely affect the results of this thesis.

The influence of disposal patterns on the variability in medieval pottery assemblages has already been alluded to above. The observation there was that the contents of a medieval archaeological assemblage do not reflect the activities of the immediately surrounding area. We must therefore postulate some intermediate stages which rubbish passed through after breakage but before burial. This has been noted by Biddle at Seacourt D.M.V. where sherds from one vessel were recovered from all the layers within a pit (Biddle, 1961-2). Similar conclusions must be drawn from attempts to reconstruct the vessels from quite large excavations where recovery of finds was good. It is rare to find all of the sherds of a vessel and quite common to find joining sherds in quite distant parts of the site. The most likely explanation is that above-ground middens existed, which were periodically removed for burial or disposal off the site. On a rural settlement this off-site disposal might give rise to so-called 'manuring scatters'. This is an explanation by field archaeologists for the presence of isolated sherds of medieval pottery in areas known to have been ploughed in the medieval period (Taylor, C. C., 1974). However, it is also possible that some of these scatters represent pre-ploughing occupation on the site, such as has been demonstrated at Goltho (Beresford, 1975, 7) and within the study region on the Frocester Court Estate by E. Price (see appendix 1).

It is suggested therefore that the disposal of rubbish in the medieval period gives rise mainly to a 'randomising' process which masks much possible variation in the composition of pottery assemblages.

The position of the site in relation to the marketing hierarchy may also be a source of variability in assemblages. If the system of distribution of goods was totally efficient then the position of the site should have had no effect on the availability of a pottery type but it is unlikely that the medieval market and fair network was so effective and sites with restricted access to markets or fairs might be expected to produce fewer imported or non-local goods, for example, than sites of equivalent status in a market town. It is a measure of the efficiency of medieval marketing that there are very few examples of this process, unless the distribution of imported pottery is interpreted in this way, so that at the coastal ports imported types were available to all but that the status of the receiving sites rises with distance from the coast.