6.7.3 East Gaulish ware among samian assemblages in Britain: frequency, supply and matters of dating

Table 18 below lists the presence of East Gaulish samian among a number of site assemblages from Britain. This is perhaps far from being an ideal sample, in so far as the representation of East Gaulish samian is established via a number of different types of measures (this being an artefact of the manner in which the samian has been reported, publication to publication). Further, with this type of data the proportion registered by East Gaulish samian will be influenced by the level of activity on the site, and its duration, during the period before East Gaulish ware became a significant import to Britain. Nonetheless, the data appear to confirm the conventional view of the distribution of this ware, specifically that it is more common in the south and east of England and at the military sites of the northern frontier. These data suggest that military sites and sites associated with the military received relatively high amounts of East Gaulish samian. This is shown by the levels at Birdoswald, Brancaster vicus, outside of the shore fort, Greta Bridge vicus deposits, South Shields and York. In other words there was a differential distribution in favour of the Roman military, presumably through the organisation of supply.

Interpreting the data presented in Table 18 is not entirely straightforward, as an important variable will be the intensity of occupation in the areas investigated which may fluctuate over time. This may relate, for instance in the case of military sites, to changes in garrisoning. To take the specific example of the County Hall site, Dorchester, Mills and Corney state that the deposition of samian sherds in this area was perhaps 'tailing off steadily throughout the 2nd century' (1993, 43), which may be significant as this would tend to result in a low proportion of East Gaulish samian, though they state that the paucity of Eastern Gaulish samian may simply be a function of geography (cf. Section 6.7.2). Further, Croom and Bidwell, in their consideration of the start of occupation at the vicus at Greta Bridge, County Durham, use the relative proportion of East Gaulish samian (as opposed to Central Gaulish ware) as an index of date, comparing the proportion to equivalent figures from other sites in eastern England with established start dates (1998, 176-8). This is a logical procedure that, in this instance, seems to generate an entirely valid guide to the approximate start date of the site. For the method to be valid, though, there has to be knowledge that other factors, such as continuous occupation, are 'equal', or at least taken into account. The figures for East Gaulish ware at sites like Caistor-on-sea, Brancaster and South Shields are so high because they begin in the later 2nd or early 3rd century, are on the North Sea littoral (hence comparatively near to the source), and besides are all related to the Roman military.

Turning to the English Midlands, reference to two groups (from Oxfordshire and Warwickshire) in the database that extend to the mid 3rd century show comparatively low numbers of East Gaulish vessels: Alcester, Baromix Factory site 1972, Phase F ('Alces3'), 7 East Gaulish vessels in a group of 67 samian vessels, and Alchester 1991, Period 6 ('Alche2') with an evidently equal infrequency. In a report upon samian from recent excavations in Leicester Dickinson notes that the proportions formed by East Gaulish ware 'are lower than normal for Britain as a whole' (Dickinson 1999d), which accords with the data collected in Table 18). She further observes a 'similar scarcity of East Gaulish products is observable at other sites in Leicester, probably because of its remoteness from points of entry' (1999d, 104). This perspective on samian at Leicester is of interest as there is an obvious contrast with the strong and wide distribution of Lezoux products across the site, for instance with the work of Maximinus i whose work overlaps the era of East Gaulish supply. It is germane to note that Leicester was a recipient of considerable quantities of Gaulish imports in the pre-Conquest period, despite its land-locked position (e.g. Clamp 1985), and also received supplies of oysters and fish during the Roman era (as attested at Causeway Lane: Connor and Buckley 1999). To paraphrase Dickinson's perspective, East Gaulish ware was not penetrating to central Britain because it was either going to the military or was being marketed in eastern England with insufficient supply to reach the midlands. This suggests a scarcity model, with sites in the east 'holding on to' samian from East Gaul.

In terms of the relative frequency of East Gaulish samian through time, the database shows a significant pattern. Table 10 shows the proportion of the total amount of samian data that each samian ware/source represents per year (drawing on the summed values of all the samples in the database, Table 10 is available here as a .xl spreadsheet allowing data to be used by the reader). The proportions formed by each East Gaulish source can be summed per year to establish the relative frequency of East Gaulish wares. This produces the following picture, to take twenty-year points through from c. AD 150: in c. AD 150 East Gaulish samian represents just 3% (2.9%) of the samian records for that year, but through the second half of the 2nd century this rises to around 7% (c. AD 170: 7.8%; c. AD 190: 7.1%) and then rises again to approximately 12% through the 3rd century (c. AD 210: 12%; c. AD 230: 9.8%; c. AD 250: 12.4%; c. AD 270: 12.1%). The source data behind these figures are drawn from a wide range of samples from across the province and from various site types, and so there is a moderating effect. The figures show, significantly, that East Gaulish ware is relatively infrequent among samian groups of 3rd century date, or which extend into the 3rd century. That the proportions are low is not surprising for two reasons. Firstly, the overall quantities of East Gaulish samian imported into Britain were not of the massive scale seen with Lezoux ware: supply was patchy and only rarely does the ware occur in any quantity in stratified groups. Secondly, it is clear that large numbers of Lezoux vessels remained in use into the 3rd century (Section 5.8), while some proportion of vessels in 3rd century groups will be residual Lezoux items. These data, therefore, appear to be a fairly reliable index of the overall frequency of East Gaulish ware in Britain.

The infrequency of East Gaulish vessels amongst assemblages from the midlands and west of Britain, as compared to the east is likely to have implications in terms of dating. This may be so because site horizons in the east of England and/or at military sites of 3rd century date may be more readily dated and dated more accurately, since they have associated East Gaulish wares, than is possible in the case of contemporary horizons from elsewhere which lack both East Gaulish samian and chronologically diagnostic coarse wares.

The Birdoswald Spur site produced a moderate sample of East Gaulish samian. An inventory of the group is reproduced here as it is constitutes an example of the diversity of East Gaulish ware occurring on sites in Britain. Approximately 22 vessels from Eastern Gaul are represented, accounting for c. 8% of all the vessels represented (among a samian assemblage of 313 sherds; a further item, either East Gaulish or from Lezoux, occurs (cf. above)). These items show a variety of sources and forms represented. Occurring through the sequence, they are generally not closely datable. As elsewhere, some are very likely to have arrived at the site during the 2nd rather than the 3rd century (see Table 18 above).

© Internet Archaeology URL: http://intarch.ac.uk/journal/issue17/1/6.7.3.html
Last updated: Mon Mar 7 2005