By Patrick Foster
Shiant Islands Project (SHIP) 2001
The second year of excavation on House Island at site HI15 (NGR NG 41999726), a blackhouse and its associated buildings, was undertaken for two weeks in July 2001. Internal levels were reduced and a complex drainage system revealed, numerous hearths, internal divisions and a possible pottery firing pit containing the large part of a hand-made pottery waster were revealed. Wall footings and layered deposits of burnt peat soil of pre-blackhouse date were also examined. The northern attached outbuilding which was found to be partially buried within a massive limpet midden excavated last year was investigated further revealing stone structures and deposits below both outbuilding and blackhouse. Similar features were found in excavations beside the external west wall of the blackhouse. Excavation was started on the small outbuilding attached to the southern side of the blackhouse which was found to possess a winnowing hole in the west wall opposite the east entrance.
To date the excavations have produced over 2,000 sherds of hand-made pottery mostly associated with around 450 sherds of early modern glazed factory products, for which David Baker (Stoke on Trent Potteries Museum) has given a spot dating of between the early mid 18th century and 1865. A small amount of grass tempered hand-made pottery was retrieved, which along with a soap stone spindle whorl and a fragment of soapstone bowl indicate a Norse occupation of the site. Several medieval rims dating to around the 14th century have been identified, and uncomfortably, a single sherd of an Iron Age vessel with applied zig-zag decoration. Final ceramic analysis will hopefully provide a link with the early medieval to the early modern period through the analysis of the hand-made material.
Of note is the assemblage of lithics, which amounts to over seventy struck flints and cores, almost all being found in the 18th century deposits. Several worn strike-a-lights and a musket gun flint are present. There are no natural deposits of flint on the islands and so far no sea borne flint pebbles have been found on the beaches. The nature of the assemblage suggests that reducing flint nodules was taking place in and around the house. There is also evidence for the working of the local basalt possibly to produce gardening implements i.e. hoe blades. At least one of the several fragments of rotary hand millstone that have been found was manufactured locally from an outcrop of coarse-grained Tertiary syenite on Mary Island.
Several other sites were investigated. Site HI 26 (NGR NG 42099736), which in the original survey was thought to have been a possible boat shaped stone setting, sitting on a skyline position on the ridge above the blackhouse, was found on excavation to be subrectangular in shape. Some of the large stones had toppled giving a false pre-excavation shape. Removal of the peat cover revealed a concentration of small stones at the southern end, which is assumed to be the remnants of a disturbed cairn. Below them a baked hard peat soil surface was encountered which was left intact for the possibility of future dating. The site was re-turfed.
Further on top of the plateau, east from HI 26, a small cairn HI 29 (NGR NG 42109743) had been built on a small hump of bedrock in the centre of a boggy depression. Pre-excavation surveying had suggested a small heel shaped cairn, but excavation revealed that the cairn was composed of only several large blocks that had been pulled apart thus giving this impression. The cairn has obviously been 'plundered¨ at some time and it is fairly certain that the cairn within HI 24 suffered similarly.
On Rough Island two sites RI 2 (NGR NG 41669825) and RI 41 (NGR NG 41179829) were evaluated with small trenches to gain dating material. RI 2 is located on the high central plateau of the island and is composed of an early modern shieling hut built out of, and on top of, an earlier circular mounded structure in which several radial walls are suggested. An Iron Age date was therefore suspected for the earlier monument. Two trenches were cut into the sides of the mound. Trench 1 on the northern edge produced a small collection of plain hand-made pottery. Dating is not immediately possible with such undiagnostic material, but the fabric, form and texture strongly suggests an Iron Age date. The excavation only removed the upper peat soil and terminated at a level of approximately 0.30m. depth where stone work began to appear in a fixed attitude and a small area of baked peat soil was encountered. Trench 2 was cut up to 0.40m. into a very grassy, humped eastern side to the mound, which was found to be entirely composed of layered deposits of limpet shell and clay. A small quantity of hand made pottery was recovered much of which is similar to the hand-made 18th century 'craggan¨ ware found at the blackhouse. The midden may therefore be related to the shieling rather than the monument below it.
Site RI 41 is located down on the southern coast of the island and is composed of an enclosure, massive stone platform and at least one circular stone built hut. The construction of a modern sheep fank with associated enclosure walling appears to have severely damaged the earlier monuments. The original survey of the site encouraged a comparison with the platformed Neolithic site at Alt Chrisal on Barra. Three trenches were cut, one across the early enclosure wall and two next to the round hut on the platform. Again only the upper peat soil was removed down to archaeological deposits, but the excavation produced abundant pottery, which is not comparable to any other found so far on the islands (except for a single sherd at RI 2). It is hand-made, thick, gritty and friable, fired at a low temperature and is undecorated. Trench 1 and 2 both revealed stone faces of internal walling belonging to two further huts. Within their compass the upper surface of collapsed walling and large flat roofing lintels were revealed. Trench 3 across the enclosure wall showed it to be 2m. wide with well built external faces. The humped profile revealed in the section suggests that originally it may possibly have attained considerable height. After a more leisurely consideration it is now thought that the platform is most likely a natural formation resulting from the collapse of the upper cliff face. The ceramics do not compare with other Neolithic pottery found in the Hebrides or with any of the material found so far on the Shiants which reduces the possible date range of the material considerably. A late Bronze Age-Early Iron Age date is not out of the question, but the well built rather massive enclosure wall may suggest an early medieval date and it is tempting to link the site with the 7th-9th century AD early Christian cross stone found in the excavations on House Island last year.