GeologyNatural HistoryHistorical ArchiveSea Room by Adam NicolsonWestminster Shiant Isles Expedition 1958 & 2002 width=Sign the Guest Book
Archaeology Icon


by Patrick Foster


The Shiant Islands lie some 6Km. off the east coast of Lewis in the Outer Hebrides of Scotland. They comprise of a group of three islands, Garbh Eilean (Rough Island), Eilean an Tighe (House Island) and Eilean Mhuire (Mary Island) and a chain of small rocky islands and stacks Galtachan. With even a cursory glance at their outline in map form one is immediately struck by their similarity to the lands of the New World. They appear to be a distorted, yet startlingly identifiable cartographic twin of Greenland, North and South America with Alaska and the Aleutian Islands, not only in shape, but in geographic order. This is not a new revelation, it is a reaction often put into print by many other observers, and is by no means the only distinctive and interesting feature of the islands. Geologically they are strikingly singular, having only a limited range of native rock types more or less restricted to dolerite, especially when compared to the diverse suite of rock types and minerals to be found on Skye only a few miles to the south across the waters of the Little Minch. The Shiants do however display one of the most extant and dramatic examples of columnar dolerite in the British Isles, which can easily be compared favourably with the splendour of Fingal's Cave on the island of Staffa whose columns are dwarfed by those of the northerly cliffs of the Shiants. The islands also boast the largest Black Rat population in the British Isle, which to some may not appear to be a worthy attribute yet it is certainly of importance to the naturalist and, perhaps surprisingly, the archaeologist. The archaeological interest is two-fold. Firstly on the Shiants the rat has filled the place usually held by the rabbit as being the greatest danger to the archaeological monuments of the islands. The monuments usually affected are those that use mainly drystone walling packed with earth in their construction since rats like nothing better than to burrow their rat-runs along and amongst the stone work eventually undermining the structure causing it to subside or tumble. Over a period of time even the most massive structure can succumb to the rodent be it rat or rabbit.

In recent years archaeological interest in so called 'marginal areas' has grown considerably giving birth to projects, conferences and publications concerning the, amongst other places, the Western Islands of Scotland. In 1987 Sheffield University began the ten year SEARCH project investigating, by excavation and survey, the archaeology of the Uists, Barra and the southern Outer Hebridean islands (Branigan and Foster 1995 and forthcoming volumes in the Sheffield Academic Press Series). A similar project to investigate the Shiant Islands was conceived when Adam Nicolson, the owner of the islands, was commissioned to write a book describing their ancient and recent social and economic history, natural history and his family's personal involvement with them. He in turn commissioned the present author, on the basis of the SEARCH project results, to undertake a general archaeological survey and to implement a series of investigative excavations on a range of the islands monuments which would be incorporated into his forthcoming publication. The work was scheduled for two weeks in late May 2000 and, although the excavation part of the work became concentrated on a single site, the overall outcome was so successful that it was decided to launch a long term archaeological campaign, expanded to include environmental research, to continue over a number of years and encompassing a much wider range of sites.


The experience gained from a decade of fieldwork in the Outer Isles was put to use and much the same methods were used for the survey. The selection of personnel for fieldwork on small uninhabited islands, especially where time and money is short, is of great importance and can have a distinct effect on the methodology used. Not only must the individuals be compatible, they must be capable of cheerfully surviving under less than ideal conditions, and most importantly be experienced enough to work unsupervised. From almost the beginning of the Hebridean experience in 1988 with Sheffield University on Barra it has been the custom for the author to encourage the participation of professionals and students from the Czech Republic. Maintaining this tradition the Shiant survey work was undertaken by author, who located and plotted the sites, with Jana Žeckytzová and Vanda Pryhouskpova (both professionals working for the Institute of Prague in Prague Castle) who produced measured field drawings for each site as well as stone-by-stone drawings of selected monuments.

The three main islands of the group were surveyed, there was no time to visit the smaller islands of the Galtachan chain, and the sites on each island were numbered from one onwards, but with the pre-fix of HI - House Island, RI - Rough Island and MI - Mary Island to distinguish between them.

With reference to Harvie-Brown's map of 1889, the first edition OS map and a number of descriptive references gathered by Adam Nicolson the sites surveyed were plotted on an OS base map. Many of the sites were already marked on the maps, but a number of sites described in the literature appear to have confused locations or were missing altogether. Sites not marked on the maps were plotted by observed estimation. Hopefully these sites will be checked by GPS methods at a later time, however the plotting is considered to have over 90% accuracy to within one meter. Although each island has its own distinctive topography and shows the surface evidence for different land uses the rules of survey remain more or less constant. These small islands, in concert with the rest of the Western Isle, fall within the upland range of landscape being wild and rugged in nature. The visibility of archaeological surface features is therefore limited to some degree because of the rough broken topography, but this is somewhat countered by the general lack of trees or large shrubs and the low grass height maintained by grazing sheep. Mary Island is also exceptionally benign with its rolling grass covered surfaces ruffled only by extensive wide lazy beds, but even here there are areas of less even landscape. The general survey methodology is to walk virtually every square metre, with constant reference to the natural features marked on the map, which are usually quite accurate since they are plotted from aerial photographs, so that ones exact position is always known. With so many small terraces and crags it is all too easy to become misplaced.

Unless a site is obviously identifiable, interpretation and especially dating is only tentatively attempted. Experience in the southern islands of the Outer Hebrides has shown that small islands naturally can be very limited in certain resources, and surprisingly, in many areas readily available stone suitable for building can be very scarce. Also as a feature of human nature it is natural that if a construction is required and suitable materials are at hand, then they will be used rather than new supplies being laboriously sought. In this way many monuments are reused, robbed or totally raised to the ground. To whichever action they are subjected to the monument is often not easily recognised afterwards.

Field walls and enclosures tend to take second place to 'site' monuments in the early stages of field survey, but they were however plotted when encountered. As a monument type they are slightly more problematical with regards to accurate plotting, especially the older ones which are often found only as remnants with only a few stones left in situ, and would benefit from GPS accuracy. Although it has been possible to date field walls elsewhere in Britain using only their typology (Wildgoose 1991) it is rarely possible in the islands. Hebridean field walls can change their construction design many times in a single unbroken length depending on and reflecting the characteristics of the country over which they are built. In pre-modern times stone quarrying for field walls was not undertaken and only those materials at hand, either field stone, turf or sods, were used alone or in combinations. If however time and labour can be focused upon the problem walls and entire enclosure systems can be dated by bringing to bear a range of archaeological techniques, including the determination of stratagraphical relationships between the wall and other monuments upon which they impinge or are impinged by. More modern field walls can usually be dated by referring to maps and the most recent enclosures are almost entirely made from post and wire fences whose remnants can usually still be found laying on the surface. If these remnants have disappeared the line of the enclosure can often still be found appearing as a line of stones laid along the wire to prevent sheep from pushing under the lower strand.

As one of the most enduring monuments in the landscape field walls may be of great antiquity and can have remained in use for almost indefinite periods of time only disappearing if their structure is physically removed or if they become buried from sight. In the landscape of the Highlands and Islands the growth of blanket bogs over the two thousand years would suggest that the last possibility could be a common occurrence.

Site Categorisation

Without the benefit of excavation, and sometimes even with it, site categorisation cannot be lightly undertaken and is generally restricted to identifying sites whose interpretation is fairly certain. This may appear to be excessively cautious, but it is considered that rather than cloud the assessment of a site with misguided interpretation and misconceptions it is better to describe the monument in terms of stone settings of a certain construction.

Archaeology Index Page