2.6.8 Sampling

Although huge areas of ground were uncovered and very large numbers of features examined, this was a rescue excavation and was constrained by the sheer scale of the operation, the availability of both a workforce and the funds required to keep them going, and by the climate which, with virtually no rainfall three seasons in a row, made conditions exceptionally difficult.

The objectives of the project were defined on the basis of spatial observation and sampling, not total excavation. A sampling policy was established at the outset of the project which provided for the recovery and recording of the evidence needed to ad dress the major questions in the project research design.

The sampling policy provided a framework in which the evidence could be classified under six broad headings - post-built structures, Grubenhäuser, pits, linear features, open spaces and surface midden deposits. In addition, work at the southern en d of the site revealed a series of complex stratigraphies which required special treatment.

Fig. 2.9 Excavation of an Anglian post-hole structure

The principal questions relating to the post-hole structures were concerned with modes of construction, layout and the use of space, dating and status. The strategy was to half excavate all post-holes, to recover a 50% sample of the potential da ting material, and to provide sufficient opportunity to examine the post-holes in section with a view to isolating structural characteristics. Only very rarely was this rate of recovery exceeded and then only if a feature had been found to be particularly productive (see also 3.6.1 Post-Hole structures).

The Grubenhäuser offered a particular problem. It is remarkable that these features, the most distinctive characteristic of Early Anglo-Saxon settlement, are so little understood. At the outset of this project we did not know what they were for, how they were constructed, nor how long they remained in use; the state of knowledge at the time was more a mass of wildly conflicting interpretations. A new terminology had been introduced, the 'SFB' or Sunken-Featured Building, but many interprete d the letters as meaning Sunken-Floored Building. The more important question, regarding the function of these structures, remained unanswered. The discoveries at West Stow in the 1960s provided the starting point for the examination of the West Heslerton 'Grubs'. Despite the recent publication of more than 200 such structures from Mucking, West Heslerton still has a key role to play in attempting to provide a more rational understanding of these distinctive features than the filthy hovels of E. T. Leeds or the weaving sheds that prevail in the more recent literature. The handful of seeds and fairly large animal bone assemblage recovered from West Stow and the one or two boxes of animal bone from Mucking represented almost the sum total of environmental e vidence from this period when the project began.

Fig.2.10 Fully excavated Middle Saxon Grubenhaus

At the outset of the project, an approach was made to Earthwatch who supported a detailed environmental pilot study. A good potential for the recovery of carbonised material in particular was identified immediately and, following that, a sampling strat egy was devised for the Grubenhäuser. These large features or 'dustbins' as they so often became, offered the greatest potential for the recovery of good assemblages from which to approach the broader landscape reconstruction. Having carefully dissec ted three Grubenhäuser with the Earthwatch teams, it was decided to empty all of the Grubenhäuser using a variety of contrasting techniques and then to float the cruciform baulks in 0.25m square blocks per identifiable layer. This generated a hu ge number of samples, the contents of which can now be examined on a spatial as well as a collective basis. If we are to understand the formation processes associated with the in-filling of these features fully, detailed evidence is essential. In addition , in order to determine whether these were sunken-floored or cavity-floored buildings, careful examination of the primary fills was necessary.

Fig. 2.11 Partially articulated animal skeleton found within the fill of a Grubenhaus

Each of the 130 Grubenhäuser examined contained a unique sealed assemblage, the recovery of which was an essential part of the excavation programme. As a result, the site has produced a large body of reliable evidence offering the greatest possibl e potential for interpretation. The extraordinary numbers of animal bones recovered are of immense importance and will surely form the basic comparative sample needed in order to set the material from York during the preceding and succeeding periods in a wider setting. Rarely have the Grubenhäuser suffered from intrusive disturbance or major residuality and thus the assemblages that they contained, although not yet precisely dated, are derived from large and effectively single period deposits in stru ctures which primarily occur during the first half of the Anglo-Saxon phase of the settlement, i.e. from c. AD 475-AD 675. In addition, a number of examples clearly belong to the Middle-Saxon period and are dated by coinage and imported ceramics pr oviding important comparative data. Some of the Grubenhäuser were clearly still in use right up to the end of the life of the settlement when they were filled in with ash/daub deposits and gradually formed natural fills. Without going into detailed a nalysis it appears that the later Grubenhäuser were generally much larger than those associated with the early phase of activity (Sse also 3.6.2 The Grubenhäuser).

There were very few pits, these being a rarity on any sites of this period. Generally they had to be more extensively excavated with up to 75% examined simply to try and identify their initial function. In most cases, it seems they were quarry p its for a particular type of soft red chalk which was preferred for the construction of ovens, furnaces and the one malt kiln that survived, with the base partially intact, beneath the headland. The material was so distinctive that in the case of the malt kiln part of the superstructure was identified in the rubbish filling of a Grubenhaüs some six metres away, where it looked both to the excavators, and to some of those who had worked at West Stow, remarkably like the features described at West Stow as hearths 'where they had collapsed into the pit beneath'; this has important implications for the interpretation of the superstructure of these features (see also 3.6.3 Other feature classes)

A small group of 'cooking-pits' was so distinctive as to deserve total excavation, particularly as there were so few (less than ten), and thus may perhaps result not from everyday cooking but from major feasts and therefore have a ritual context .

Fig. 2.12 A multiphase sequence of linear features excavated in 1995

Linear features were conspicuous by their absence from most of the northern half of the site; generally it was agreed that no more than 25% of the linear features in open space would need to be examined in order to characterise the features and in particular to date them. At the southern end of the site where many stratigraphic relationships survived, the linear features required much more intensive excavation, a situation exacerbated by the gradual realisation that the southern third of the sit e appeared to lie both on top of a late Roman prototype and within a sequence continuing into the late 8th or early 9th century. The need to tie down this remarkable and quite unanticipated settlement sequence meant that efforts were concentrated in one a rea at the southern end of the site, specifically 11AD, BD, and CD, which had been opened in 1990 and was then excavated in 1991 and 1992. The area immediately opposite, 11BA, was on the other hand less intensively examined once the general plan and major features had been identified and sampled. The almost complete lack of any property boundary features, whether stake fences, ditches or gullies in the northern half of the settlement is remarkable and contrasts dramatically with the contemporary picture o n the continent where individual house plots arranged along trackways are a distinctive feature. The enclosures at the southern end of the site, which appear to have been maintained and redefined throughout the life of the settlement, do not reflect the c ontinental plan and seem to derive mostly from the established Late Roman pattern. In Site 12 the complex of enclosure ditches spanned the life of the settlement and required almost total excavation to extract the full stratigraphic sequence. The importan ce of this sequence of enclosures cannot be underestimated, since the material derived may provide the key required to isolate the Early and Middle-Saxon assemblages. It will be important to establish the extent to which the enclosure structure was define d in the Roman period and to determine to what extent they played a role in the sub-Roman phase, where a particularly crudely made series of 'Roman' ceramics appear to demonstrate activity into the middle of the 5th century AD (see also 3.6.4 Layout of the settlement, property boundaries and enclosures).

The buried soil deposits and open area spreads of debris offer important potential for understanding the development of the settlement. 3D recovery of all finds from these deposits, which tended to survive in isolated patches between plou gh-damaged zones, was felt to be a priority. The recovered resource thus offers great potential for the interpretation of disposal practices and fragmentation as part of the site formation process. A number of areas where the soil profile was remarkably w ell preserved produced almost no cultural debris, indicating that the settlement was well maintained. The housing areas in particular were kept very clean; elsewhere, the casual viewer may be forgiven for thinking that the site looked like a butcher's yar d, albeit one spanning over two hundred years of build-up. In parts of 2DA this is the only sensible conclusion that can be offered for the sheer quantity of material recovered from a number of Grubenhäuser and associated pits. The large quantity of three-dimensionally recorded material offers great potential not only for direct analysis but also for retrospective simulated sampling experiments, which could provide a testable environment within which to view the strengths and weaknesses of different approaches to artefact recovery.

Sampling policy for the remainder of the deposits was determined on a feature by feature basis, with the general objective to examine 50% where possible. The largest single feature, the relict stream channel, proved to be so complex and extensiv e that it was only possible to examine this through a series of five sample trenches. These indicated that the stream channel contains a huge quantity of additional animal bone, much of it mineralised and probably prehistoric in origin (see also 3.6.5 Spatial patterning and midden deposits).


© Internet Archaeology URL:
Last updated: Tue Dec 15 1998