13.1.2 Guide to the 'Normal Patterns' per site type identified by the project, with which specific site assemblages and groups may be compared

• Samian occurs at virtually all investigated Roman military sites.
• Samian is comparatively frequent among site groups at military sites (both legionary and auxiliary) in Britain in terms of the proportion of the overall quantity of pottery present.
• Interventions at military sites often yield large assemblages of samian, with the exception of 'campaign' forts of the early Conquest period and in Scotland; Trajanic and Hadrianic groups of any size are, notably, uncommon.
• Attention to the ratio of decorated to plain samian forms at sites shows that military sites have, comparatively, a high incidence of decorated vessels (bowls). The army were major consumers of decorated bowls to which they evidently had reasonable access. Soldiers of officer rank may have been the principal users of these bowls.
• There is evidence from the incidence of samian ware to support the theory that the supply to military sites was organised in zones.
• Military sites/units appear to have received comparatively regular consignments of samian. There was probably a greater turnover of samian at military sites than at civil centres and so samian assemblages from civil sites may appear 'older' when compared with assemblages from contemporary military sites. In so far as dating of samian is influenced by its occurrence within military horizons, implications for dating frameworks arise.
• A number of large groups of samian have been encountered at military sites, particularly of late 1st century date. These groups have been thought to represent dumps of broken or unwanted stock. While this may be the case, a possible alternative explanation or aspect is that these deposits have a symbolic dimension representing termination deposits marking (with culturally valued material) endings in site phases or occupation.
• Spatial analysis of samian at military sites shows potential for elucidating differing functional and status zones. Attention to the context and distribution of samian inkwells at military sites (Willis forthcoming a) reveals a close correspondence with headquarters buildings (principia) and the commandant's accommodation (praetorium). Given our knowledge of the likely identity of buildings and the status of their occupants, and of the overall organisation of forts and fortresses, future spatial studies of samian in these domains should prove informative.

• Samian occurs at extra-mural sites (vici, canabae) outside Roman military installations.
• These sites appear to have consistently high proportions of samian among their pottery assemblages when compared with other types of site in Britain.
• Vici and canabae show high ratios of decorated to plain samian forms, and overall have proportionally more decorated vessels than any other site type, including forts and fortresses, by some margin. This difference reflects their likely identity as places with Romanised populations, associated with trade, recreation and entertainments for garrisons and travellers. It is also possible, if not likely, that soldiers, especially officers maintained homes in these quarters for their families and perhaps as a refuge from the fort. In such milieux the acquisition of decorated bowls, possibly as drinking vessels in some instances, is explicable.

• Samian occurs at all major civil centres, where it is a comparatively frequent find among site groups. Multiple factors account for the comparatively high levels of samian at these sites.
• Interventions at major civil centres typically generate large assemblages of samian, often from deeply stratified sequences which the samian can help to date.
• Major civil centres tend to have lower proportions of decorated samian (average 25% decorated) than at both military sites and extra-mural sites outside forts/fortresses.
• Stratified samian groups and assemblages from major civil centres tend to have significantly more 'older' vessels than do military sites. This is perhaps partly due to residuality, but almost certainly reflects a slower turnover than at military sites. This contrast may vary by degree over time. Much 'old' samian was evidently in use even in London during the early 2nd century.
• As more samples form major civil centres become available for study the scope for identifying and mapping differing functional and status zones will be enhanced.

• Samian is present at all investigated smaller civil centres of the Roman era but quantitative information confirms that it is an infrequent find. In fact it is no more frequent than at rural sites, implying these sites had a similar relationship to samian ware.
• The old tacit model suggesting that communities living in smaller civil centres were less interested in samian as a means of cultural display is challenged by the fact that samian is prominent in burials and structured deposits associated with smaller civil centres. Communities at smaller civil centres evidently treated samian differently from other pottery types and presumably valued this ware.
• Smaller civil centres might be expected to have had easier access to samian than rural sites in so far as they were located on major routes along which supplies of samian and traders will have moved.
• In the round the scale of smaller civil centres is such that some extensively excavated sites have yielded very large assemblages of samian.
• Curation of samian vessels is likely; it may have been used with care and/or occasionally.
• Proportions formed by decorated samian vessels are comparatively low among these sites, in the round lower in fact than is the case at rural farmstead sites, which probably relates to the fact that some rural sites had wealthy inhabitants of relatively high status compared with smaller civil centres.
• Some spatial analysis of samian occurrence at smaller civil centres has been undertaken which shows potential for characterising building and room function and status.

• Only a modest number of samples of samian from formal religious sites are available for analysis within the literature. Samian is generally not particularly prominent among the assemblages of pottery from these sites, and at a number of sites forms very low proportions. There is a contrast, though, between this impression in the case of formal/monumentalised religious sites and the use of samian in likely ritual deposits in shafts, wells and other water places and in structured deposits at settlement sites. In these latter cases samian appears as a frequently selected artefact, often in some quantity.
• Samian is a frequent inclusion among grave assemblages, as it is on the continent (cf. Section 9; e.g. at Nijmegen: Stuart 1977). In Britain samian occurs more frequently in burials associated with smaller civil centres and rural communities than it does in burials associated with major civil centres. There was a consistent grammar to the samian vessel types deemed appropriate for inclusion in graves which extended through time and geographically. Decorated forms (particularly bowls) are virtually always absent from burials, independent of the type of site with which the burial/cemetery is associated. Seemingly larger samian forms generally were also avoided as grave goods. Samian vessels 'with a biography' were particularly selected as appropriate for inclusion in graves, that is repaired, worn or old vessels, etc.

• Knowledge of samian collections from villas is surprisingly limited, and the characterisation of trends is accordingly hindered. Few villas have been published with full lists of associated samian and/or quantitative data to modern standards. Proportionally less interventionist excavation has been undertaken at these sites in recent years and so the number of samples reported using contemporary methods is and will be limited. Villa sites published latterly have, for whatever reason, tended to be later in date and therefore not contemporary with samian.
• Many villas have samian assemblages which are of modest size and/or proportion. (This can be seen at Latimer in the Chess valley: by sherd count samian accounts for 6.7% of the pottery during villa period 1 dating to c. AD 150-200, and 1.2% during villa period 2, with sherd totals for these periods (all pottery fabrics) being 490 and 877 respectively; however, the 33 samian sherds from period 1 come from just six vessels. Hence Detsicas stated, 'It will be seen ... that samian ware never formed a very notable part of the pottery assemblage at Latimer' (1971, 132).
• Some, though not all, rural sites that later develop into villas have high proportions of decorated samian in their early phases.
• Proportions of decorated samian as opposed to plain forms are generally low among the samples from villas available to this study. More samples are required to verify the degree to which this may be a general trend. It is possible that in elite households the place of decorated samian was taken by more ostentatious materials for display at tables, such as glass and metal containers, perhaps including silver.
• Otherwise rare and rather exotic samian ware types occasionally appear at villas, but not with the frequency that might be predicted. Such items include a samian candlestick and a samian lamp, but these are exceptional.

• Samian occurs at virtually all investigated rural sites occupied in the Roman era.
• Samian is comparatively infrequent among site groups at rural sites in terms of the proportion of the overall quantity of pottery present.
• Some rural sites have yielded large assemblages of samian.
• As at smaller civil centres, curation of samian vessels is likely; it may have been used with care and/or occasionally.
• In the Conquest period rural sites often acquire small samian assemblages but with a high incidence of decorated forms present.
• Proportions formed by decorated samian vessels are, however, comparatively low in the 2nd and 3rd centuries.
• Burials and structured deposits related to these sites regularly include samian, while there is the intermittent occurrence of 'exotic' types at these sites: these trends demonstrate that samian was valued amongst rural communities.
• The value of samian as a guide to dating is diminished at rural sites by the character of site formation processes, and the types of features in which pottery occurs, but samian is still often vital as a dating index in constructing broad chronologies for these sites.
• Spatial analysis of samian at rural sites shows potential for elucidating differing functional and status zones.

© Internet Archaeology URL: http://intarch.ac.uk/journal/issue17/1/13.1.2.html
Last updated: Mon Mar 7 2005