PREVIOUS   NEXT   CONTENTS   HOME 

3.3. Period 3, c. AD 20 - 55

Assemblage composition (see Tables 3, 4, 5, 6, 7, 13 & 14)

The transitional period was characterised by further significant change to the settlement, including the creation of a road network and temple precinct in the central zone (Atkinson 2002). The composite assemblage displayed some continuity with the previous period, with small decreases in table wares and drinking vessels matched by small increases in jars and preparation/serving vessels. Whereas most of the individual form types were present in similar proportions to the previous phase, the most notable changes were in the imported vessel classes. Significant decreases in wine amphorae (Table 5) and Gallo-Belgic imports were only partially compensated for by the arrival of very small quantities of samian (bowls, cups, dishes and platters). The decline of Gallo-Belgic imports at Elms Farm coincides with the floruit of imports at Sheepen, Colchester (Pitts 2005c; Pitts and Perring 2006), and could be a result of shifting power relations and trade towards the Camulodunum oppidum in the early to mid-1st century AD. The bulk of the assemblages were from pits, with a higher proportion of ditch assemblages than previous phases. Jar forms comprised the highest proportion by functional class in all features apart from ditch 11062 in area N, which was dominated by dining vessels.

Correspondence analysis: excavated area (see Figures 4a and 4b)

CA of pottery deposition by excavated area revealed a continued pattern of the bulk of imported pottery and drinking vessels corresponding to excavated areas from the southern zone, plotted in a continuum corresponding to areas M and K (and G) in the top-left quadrant, to N and L in the bottom-right quadrant. The presence of area G in this group probably indicates a secondary centre of high-status consumption in the north-west corner of the site. Otherwise, beyond the small quantities of samian ware corresponding to features from the northern zone of the site, there is little patterning of obvious note.

Correspondence analysis: assemblage/feature (see Figures 11a and 11b)

Once the outlying forms and features had been removed (see Table 24), several trends were apparent. First, there appeared to be a much higher level of homogeneity in assemblage content across the site, with no apparent clustering of features by zone, as seen in the previous phases. This suggests that consumption habits had become less hierarchical, corresponding to the rapid decline in wine importation and consumption in this period (which previously had been focused in the southern zone of the site).

The second trend of note was the clustering of vessels in Gallo-Belgic fabrics (beakers, butt-beakers and platters) and forms in local fabrics (cups, butt-beakers and platters), corresponding to a series of assemblages from both northern and southern zones. It should be stressed that such patterning was by no means an isolated occurrence at Elms Farm, with several further examples occurring in contemporary assemblages from the region at Sheepen (Colchester) (Pitts and Perring 2006), Baldock and Braughing (both in Hertfordshire) (Pitts 2005c). These results seem to reflect the deliberate deposition (and possible use) of meaningful suites of vessels with similar functions or fabrics as culturally specific dining and/or drinking sets. The principal evidence for this comes from contemporary transitional period funerary assemblages, which featured similar combinations of complete Gallo-Belgic imported vessels. This trend is especially notable in the middle- to high-status graves of Birchanger, Little Waltham, Southend, Stansted Airport (Havis and Brooks 2004), and Stanway (Crummy 1992), where full suites of Gallo-Belgic imports are apparent (Pitts 2005c). Although such a phenomenon could be explained as a product of assemblage size, with larger assemblages more likely to have a greater range of vessel types in the same fabric or style, the assemblages in question were of variable size (from 1-6 EVEs in this case). Moreover, similar patterning at other sites (particularly Sheepen in the same period) involved a greater number of much smaller assemblages, reducing the probability that the contextual associations identified through CA had occurred at random. A similar cluster of dining vessels can be seen immediately adjacent on the right-hand side of the plot, although this pattern appears much less significant, corresponding to fewer features (including ditch 11062, already noted for its high level of dining vessels).

 PREVIOUS   NEXT   CONTENTS   HOME 

© Internet Archaeology URL: http://intarch.ac.uk/journal/issue21/2/3_3.html
Last updated: Tue May 08 2007