Cardiff Studies in Archaeology - Specialist Report Number 18

DISCUSSION

next page

This year excavations have furthered our knowledge of the settlement and together with the information accruing from post-excavation analysis are beginning to provide a detailed understanding of the site and its significance.

Radiocarbon dates have confirmed the first millennium date of the settlement remains on mound 1 and have placed it in the period between the Middle Iron Age and Late Iron Age (Barrett and Foster 1991). This is a period of some interest as it marks a profound change in the material culture and architectural complexity of the Atlantic Province. There is a movement away from the monumental architecture of brochs and wheelhouses towards the very understated nature of the cellular structures of the Late Iron Age. Coinciding with this change we see a marked decline in the quality of pottery and in the complexity of decoration on vessels. In contrast to this decline in the quality of pottery there is an increase in the quantity of artefacts and in particular of personal artefacts such as pins, combs, brooches and beads. These become increasingly elaborately decorated in the Late Iron Age.

The Bornais settlement was relatively rich in artefacts but it does not contain the exotic imports and technically complex metalworking that we see in some Late Iron Age settlements. The pottery is not elaborately decorated but it is still well fired. Architecturally the building seems to have originally been a wheelhouse but it was at least partially converted after it was burnt down. The end of this occupation was marked by the systematic demolition of the structure and its deliberate infilling. This mound was then abandoned and mound 2 became the focus of settlement activity. This abandonment seems to be a recurrent feature of settlements of this type and can be paralleled at several other sites in the Uists (Sharples forth). It contrasts with the evidence for continuity seen on broch sites such as Dun Vulan.

The excavation of mound 2 is crucial to the understanding of the settlement in the Norse period. Though it can not yet be proven it seems likely that this is the location of a settlement occupied when the Vikings arrived in the Western Isles. This settlement became the location of a Norse farm. The earliest building is a substantial timber hall as yet only very superficially explored. This was succeeded by at least two structures with substantial revetted walls. The initial rebuild was a building 19.3 m by 5.8 m and the subsequent rebuild was 12 m by 5.4 m. These three houses are by far the most substantial houses known from both Kilphedir and Bornais and they are comparable to the size of the house at Drimore. The quality of the stone work in both houses suggests access to the best stone resources (probably ancient dwellings). This suggests the settlement is the location for a family of considerable importance and that this status lasts for several generations.

It is not clear exactly how long each of the three houses lasts nor do we know the length of time between the abandonment of each house and the construction of the next house. The chronology is anchored by the material culture and radiocarbon dates from the floor of the second building. These clearly indicate a date in the eleventh century AD. There seems to be a relatively short period between the timber house and this house, as the original structure has to be covered by dumping sand. In contrast there is clearly a long period of modification, quarrying and infilling between the two later stone revetted houses. It is difficult to envisage any of the structures in this period as a substantial house compared to the three principal houses.

This sequence may indicate the relative fortunes of the family if each house is associated with an individual. The short period between the first phases may indicate a smooth succession from one generation to the other with status and position within the community being retained. In contrast the period of confused construction and quarrying between the second and third house may indicate difficulties within the family. A problematic succession to power may also be indicated by the excessive quantities of artefacts found on the final floor of house 2. This is not a casual or accidental loss of objects. Most of these finds were complete and in good condition when deposited and it beggars belief to assume they were not consciously discarded as part of the closing acts of this house. I would currently explain this deposition as part of the burial rites of the occupant. The unusual nature of this depositional rite (house floors are normally very clean) suggests the circumstances surrounding the death were unusual and that we are dealing with some sort inauspicious death. The destruction of the occupant's artefacts suggests they were surrounded by some form of pollution, which prohibited any further use of these items. This pollution and the possibility of an inauspicious death may explain the prolonged gap separating the construction of houses 2 and 3. If the family lost its status after the death of the occupant an alternative possibility is that they initially chose to avoid the position of this house when they constructed their new house. Another house is present immediately to the north of the trench.

There is as yet no evidence that any of the other settlement mounds were occupied as early as the tenth to eleventh century AD and this year excavation of mound 2A revealed this area was under cultivation in the early Norse period. Both of the long extension trenches came down onto a bone rich soil horizon, which overlay plough marks cutting into clean sand. The artefactual material from both mound 2A and 3 suggest activity at these mounds began sometime in the twelfth century or later.

On mound 3 the evidence (summarised in the 1997 and 1999 interim's) suggests the development of a farm carrying out fairly basic agricultural activities. The carbonised plant assemblages indicate a range of species were cultivated including barley, oats, flax and rye. The latter is an unusual discovery though it is not unknown. The occupants of the house also had access to large quantities of herring and these were being consumed inside the house along with cod heads. There was no evidence for processing the fish on this mound but as the middens associated with these structures were not excavated this may not be significant.

The evidence from mound 2 is a complete contrast to this picture of subsistence activity. The initial occupation of this mound is connected with some form of high temperature industrial activity, which has produced large quantities of slag. It is possible this represents iron smithing but further work is required. This activity continued for some time and is associated with a structure, which lay just outside the area excavated this year. These deposits are then sealed by middens and structures, which were rich in artefactual and ecofactual material. The midden was extremely rich in fish bones and superficial examination has noted a large quantity of jaw bones that might indicate this was a processing rather than a consumption site. The dominant artefactual material is the waste from bone comb and gaming piece production and it is clear that production was taking place inside the house excavated in 1999. The evidence suggests this mound was continually associated with craft activities and other messy acts of processing but that these changed from time to time.

The presence of these craft activities is a very unusual feature and combined with the size of the settlement make this site unique in the Western Isles and with very few parallels from elsewhere in Britain and Ireland. These unusual features appear to develop in the late Norse period and it is possible that they might be related to the development of commercial fishing for herring. This was a surprising discovery and contrasts markedly with the existing pre-Norse fishing strategy and the strategies adopted in the Northern Isles. It may have been stimulated by the close political contacts that existed between the Hebrides and Man and therefore with other areas of the Irish Sea. The urban centres that existed in Ireland would depend on outside resources for their basic survival and it is possible they were importing basic staples from areas as far afield as the Hebrides. The artefactual evidence from Bornais suggests contact with these areas and the cu alloy pins are very likely to come from Dublin. A range of other objects, including pottery, steatite, glass beads and walrus ivory, indicate the importance of trade to the inhabitants. The ability to undertake craft production would depend on these trade relationships both to acquire raw materials and to dispose of the finished artefacts and these contacts might have originated through the supply of a basic staple such as herring.

Successful commercial herring fishing depended on a reasonable sizeable community with access to a wide range of resources. Boats and nets would be required for fishing and these could have been imported or produced on the Western Isles. Even if imported they would need to be maintained and this would require access to wood and iron nails. Processing the catch is as important and would require containers such as barrels and a preservative such as salt. All of this activity would be an important stimulus to the economy of the site and the region and might explain the relative size and wealth of the community.

next page