Internet Archaeol. 17. Willis. The rarity of samian at some formal/monumental religious sites?

12.2.2 The rarity of samian at some formal/monumental religious sites?

While the sites documented above show samian to be present in and around a variety of formal religious sites there are cases where samian seems to be poorly represented at such sites. Tables 23, 24, 27 and 28 show that samian was a minor component of groups from such sites within the quantified sample available. The specific groups, where weight is the method of quantification, come from Chanctonbury Ring, West Sussex (three groups), Chelmsford, Essex (two groups) and Uley, Gloucestershire (one group); at best samian comprises only 3% of the pottery forming these groups (Section 7.2.6). When EVE is the measure, data are available for only three groups, from Chelmsford and Uley (Tables 27 and 28), among which samian forms between 2.4 and 5.7% of the pottery forming the groups. A similar pattern of infrequency is apparent vis-à-vis the temples at Harlow, Essex, and Lancing Down, West Sussex, as well as at, seemingly, Bath. A low occurrence of samian at formal religious sites is suggested by the following cases.

Harlow Temple Pre-Claudian samian/sigillata is absent from the temple site at Harlow during its pre-Conquest phase, despite the occurrence of such imported fine ware at contemporary sites in the vicinity (France and Gobel 1985; cf. Rodwell 1985, 104).
Uley Shrine At Uley the absolute quantity of samian recovered from the works of 1977-9 is low (Johns 1993).
Leicester, possible Temple at St Nicholas Circle It was noted in Section 7.2.4 that among a Hadrianic-early Antonine pottery group from St Nicholas Circle, Leicester, samian was conspicuously infrequent for such a major centre. The specific site uncovered a likely public building, thought to be a temple.
Bath Bath was a unique centre in Roman Britain, with many monumental buildings. Archaeological work at the temple baths complex (Cunliffe and Davenport 1985) yielded relatively little samian ware, forming just c. 1% of the Roman pottery assemblage (Green and Young 1985, 145-6). In truth this low proportion may relate in part to the mid to late Roman date of the assemblage examined by Green and Young. On the other hand, Bath has yielded a quite exceptional decorated samian bowl, approximately of Drag. 29 form and hence likely to be of 1st century date (Haverfield 1906, 287, fig. 55). This piece is decorated with continuous vertical 'reeding' with a band of repeated motifs depicting a helmeted head, immediately above the angle of carination. This vessel came from the Baths, known also for their votive assemblage. Sites with a special status and function, as with Bath, might be expected to yield unusual artefacts such as this bowl.

12.2.3 Samian at formal/monumental religious sites: discussion

The cases listed above in Sections 12.2.1 and 12.2.2 suggest that at some formal religious sites samian is prominent while at others it is less so. In other words, the emerging picture shows variation. A varied pattern is perhaps not surprising, though more quantitative data from sites of this type will likely assist in clarifying trends. The trend, as far as it may be discerned from present quantitative data, is for samian to be present in modest or low frequency. Nonetheless some interesting cases of presence occur. That samian is routinely present at these sites is itself a point of note, since it demonstrates that it was not unacceptable within these cultural environs, or otherwise segregated. Since it is likely that different deities will have been venerated and evoked by differing acts and procedures (cf. the Carrawburgh mithraeum) it may be that variation in material evidence from place to place is to be anticipated.

Of note, the early Roman stratified pottery group from the shrine at Uley, the three 2nd century samples from the site of the temple at Chanctonbury Ring and the two groups from the religious complex on the outskirts of Chelmsford (site K), have proportions of samian closely similar to those of the rural groups and the groups from the smaller civil centres (Tables 23, 24, 27 and 28). This trend is seen too among the material from the temple at Lancing Down, West Sussex (Bedwin 1981) for which sherd count data have been published (Rudling 1981), and seemingly at the temple at Harlow, Essex (France and Gobel 1985). This correspondence with the percentages of samian occurring at the rural sites and smaller centres may in these cases be considered appropriate since these particular temple sites, namely Uley, Chanctonbury Ring and Lancing Down were located within rural environs, while the site at Chelmsford lies on the periphery of this small town.