Artefacts...
Below are a selection of artefacts recovered form the site during the excavation:
NB. these finds relate to various phases of activity at the site and not necessarily the wheelhouse.
(click on images to launch viewer)
NB. these finds relate to various phases of activity at the site and not necessarily the wheelhouse.
(click on images to launch viewer)
Pottery...
The Iron Age of the Western Isles, in contrast to much of mainland Scotland, boasts a very rich ceramic record. The study of this pottery in the Western Isles, a significant portion of which was produced from early inadequate excavations, has not been subject to the vigorous debate that the structural sequence has seen in recent years (see for example Armit 1997, MacKie 1997, Parker Pearson et al 1996, 1999, Gilmour 2000). This trend is particularly evident in a recent publication of the excavations at Balesare and Hornish Point (Barber 2004), where pottery is considered, whether intentionally or not, as peripheral to the structural record. In addressing why this may have occurred, two main reasons are offered here. Firstly, Iron Age structural remains in the Western Isles are very impressive and remain conspicuous in the landscape. It would not be optimistic to propose that every significant islet or patch of arable land could contain some remnants of prehistoric intervention. The other contributing factor to an alienation of pottery studies in the Western Isles has been the pessimism generated by previous pottery studies all of which, from Beveridge in 1911 through to the present, stress that the uniformity in fabrics from prehistoric to the recent ceramic record hinders any detailed examination of production centres and exchange. Although criticisms can be made of Topping’s methodologies (see Lane 1990) when seeking to resolve the question of ceramic development in the Western Isles, the conclusions of his work have become the recent starting point for current research:
…without the uniformity of commercial or specialist production the relevance of classification may be limited, with patterns within the data being too ephemeral or too variable for secure archaeological identification (Topping 1985, 82-83).
Such pessimistic results from Topping’s study were arguably the result of a failure to classify the assemblage with a macro examination of the inclusions before proceeding with the neutron activation analysis (Lane 1990, 116). Lane gives examples of how the neutron activation analysis is flawed, arguing that if approached correctly (e.g. Lane 1983) patterns are visible in the pottery data (Lane 1990). Other recent approaches are seeking to further the cause of pottery studies, with Johnson for example, examining such questions as: Why should pottery change at all? If it forms its function adequately, then why change it? It is inescapable that pottery has a social role as well as a utilitarian role, and so establishing why pottery changes involves looking at the changing roles of pottery in a symbolic or social domain as well as its functional one (Johnson 2004, 2). Currently we are facing a situation whereby it is desirable for more rigorous treatment of primary archaeological data utilising modern techniques in conjunction with some fundamental reassessment of how we view the role of pottery within prehistoric societies. Recently, attempts have been made to assess pottery sequences independently of the structural sequence (see Barber 2004, 126). The intention of this approach is to see if changes or patterns in the assemblage relate to changes in the nature of the settlement and can be linked to a broader change in society, or whether ceramic developments occurred independently, or at a different pace, to settlement adaptation. The author’s intention for the pottery recovered from Bagh nam Feadag was to examine variations in the fabric, however, it soon became apparent that little difference could be identified within the collection as the majority derived from mineral rich Lewisian gneiss clays and cannot be accurately sourced (see Topping 1987). Problems in differentiating between Iron Age and post medieval fabrics have been stressed in virtually all studies of pottery from the Western Isles, and the Bagh nam Feadag assemblage is no exception. The changes that can be seen in the ceramic fabrics from Bagh nam Feadag possibly say more about changes in firing technology rather than a change in clay source or importation from elsewhere. It is proposed here that the production technique utilising reducing atmospheres in the later examples have helped to produce a uniformly harder and grey or brown coloured pottery as opposed to the softer buff and orange fabrics caused by firing in an open fire. A consistent variation within the assemblage was the surface finish which ranged from rough, unfinished, very coarse wiping or combing of the surface, to smoothed outer and inner faces, although not always occurring together.
The Bagh nam Feadag Ceramics...
A total of 2309 sherds were recovered from the excavations representing a minimum of 130 vessels. The collection weighted 36.76Kg and the diagnostic sherds consisted of 107 rims sherds, 72 bases and 54 decorated sherds. At Sollas A/B some 3000 sherds were recovered, representing a minimum of 205 vessels (Campbell 1991, 148), and at Clettraval, Scott recovered in excess of 3000 sherds (Scott 1948, 56). The volume recovered at Bagh nam Feadag is comparable with the quantities recovered from other similar wheelhouse sites although the methods of excavation and limits of scope, particularly with reference to primary floor deposits, would account for the lower end of the expected amount recovered. Also, arguably, the multiple phases of occupation at Bagh nam Feadag would suggest that a significant amount still resides at the site or was discarded during the excavation. The latter would seem unlikely as the contents of the assemblage would suggest that whatever was found was collected and retained. The only real question of how much was discarded would focus upon the techniques used to excavate and the ability to recognise material culture. As voiced in chapter one, the artefacts were recovered without any systematic recovery procedures in place and where context was assigned to a single item or bag of items it did not enable a detailed examination of the stratigraphic relationships.
Fabric...
All the Bagh nam Feadag pottery with the exception of the East Coast Gritty Ware was produced from variable, coarse, local fabrics, with the mineral components deriving from the local rock type. The only media variation is possibly the medieval sherds of a pinkish colour, which tended to only have quartz inclusions, however, these sherds may also have been made from a local clay. The bulk of the pottery assemblage consisted of undiagnostic body sherds. There were relatively fewer decorated rim or base sherds, although some were fairly substantial. With the outline given above, classification of the fabrics poses considerable problems; the main difficulty being separation of the types into distinct fabric groups, and relating these groups to form and decoration, which is the standard procedure for most ceramic studies. The slight variations that can be seen in the fabrics may not be as a result of any dramatic change in origins. The one consistent variation is the hardness of the pottery, ranging from very hard to soft, powdery or malleable. Of course, different processing styles, where clay is more refined in some examples when compared with other contemporary vessels, may well be indicative of origin or it may simply be as a result of a local event such as the availability of a purer clay, an ability to spend more time on refining the clay or the urgency with which the finished product was required. A general trend that is visible in the Bagh nam Feadag assemblage, and at other Hebridean sites type, is that the later plain style pottery tends to be slightly better fired and thus a more robust fabric (Parker Pearson et al 2004, 116). Although no grass tempering was noted or the deliberate addition of minerals, some vessels had discrete areas of inclusions, normally concentrated in the base or basal area. Further examination of this feature would be required, as well as the study of more intact specimens, to establish to what extent this was a deliberate practice or simply the result of not smoothing an area which would have been camouflaged by the effect of the fire during use. One other variation in the fabric was one sherd recovered from structure I which had a relatively grit free, uniform fabric but with large quartz inclusions, presumably to help absorb thermal shock during firing.