1 Environmental and Research Background | 2 Post-medieval Buildings | 3 Earlier Vernacular Buildings | 4 Conclusions and Discussion
In his introduction to the vernacular architecture of Britain, Brunskill (1992, 21) uses the highland cottage as the principal example of the vernacular; the post-medieval buildings of the Outer Hebrides are perhaps the most long-lived and definitive of these; some were in use until the 1970s (Burgess 1995, 24). They are relatively well understood, due partly to antiquarian descriptions and illustrations dating from the late 17th century onwards, and partly from the small number still extant, limited numbers of which have been analysed in detail. The forms of these buildings are numerous and focus around principal functions of the pastoral community: either as permanently occupied buildings in the winterton (usually in the form of longhouses); or as temporarily occupied buildings known as shielings, bothies or bothan.
These two principal types of building longhouses and shielings differ in shape, scale and quality of construction method. Shielings tended to be smaller, are often sub-circular in plan with irregular stone lintel or timber roofs, covered with turf and sometimes thatch. In contrast, the longhouses tend to be sub-rectangular to rectangular in plan, larger in size and have simple collar truss roofs of timber, overlain with a combination of thatch and turf. Shielings were not subject to the same rapid sweep of modernisation in the late 19th and early 20th centuries as the longhouses, and similar buildings are still very occasionally used for fishing and peat collection in the Western Isles. This section outlines the contemporary understanding of the nature of post-medieval vernacular buildings, drawing on written and oral sources from the 19th century to the present day. The common building materials of the Western Isles were those close to hand the stones from field and beach, the turf, clay, and peat from the surrounding landscape, and the timber from the bog or washed up on the shore.
Until the early 20th century, stone tended to be locally procured and was generally unaltered from its natural state. This reflected the intractable geological nature of the gneiss stone, and the development of skill and experience in choosing appropriate stones, rather than a lack of skill in stone-masonry. Stone would commonly be collected from fields, beaches, river beds, and scree slopes, while quarrying in the 19th century tended to be for industrial purposes.
Turfs were often used for the underlay in roofing and for capping exposed walls. Turfs or sgrothan were cut using a flaughter spade, breast plough, or an ordinary delving spade (Walker and McGregor 1996, 9) and a 'good man' could cut approximately 200 turfs in a day, with around 1000 needed for the roof of a common house (Souness 1985, 87). These tools can produce a turf that is thick in the centre and tapers to each side, which helped reduce the thickness of the lapped joints on the roof. The labour of transport would often be undertaken by all members of the community, including both women and children. Such was the ease of sourcing and building with turf that seasonal lobster fisherman in early 20th-century Lewis would often build a new bothy each winter (Malcolm MacSween, cited in School of Scottish Studies 1972). The longevity of the turfs would depend on the condition of the thatch; those on a building at Borve, on the island of Berneray near Harris, had not been replaced in 40 years (Mechan and Walker 1989, 25).
Confusion is commonplace over the definition between peat and turf and other terms such as divots, feal/fale etc. In this work, I shall follow the definition from the Oxford English Dictionary website:
'In many districts turfs are distinguished from peats, as being pared from the surface, containing roots of grass and recent herbage, and being lighter coloured, while peats are usually dug from the "moss" or bog, and consist chiefly of long-decayed and compressed vegetable matter, black or dark brown, formed from Sphagnum or other mosses.' http://dictionary.oed.com/
At Arnol, heather turf was used for the roof, whereas grass turf was commonly used in the surrounding communities (Walker and McGregor 1996, 9). The natural occurring clay at the base of the peat bogs was often used for waterproofing and providing strong foundations. Clay procured in this way continues to be used by National Trust path builders in remote locations in Scotland (John Quinn pers. comm. July 2004).
These locally procured materials were supplemented by a timber supply largely reliant on the conditions at sea rather than those on the land. The roofing and furniture requirements in the historic period appear to have been dependent largely on driftwood from shipwrecks and natural processes, brought onto the Hebridean coastline by the strong seas and prevailing winds (MacGregor 1949, 202; Fenton 1995; Walker et al. 1996, 23).
Figure 7: Derelict building at Sorisdale, Coll. Possibly the last thatched
cottage on Coll and inhabited until the early 1990s. The roof structure is of
irregular driftwood and reused timber pegged or nailed together, covered with
turf and thatch and held by chicken wire. Photo by author
Specific references to timber procurement are few but the buildings that have been studied in detail, such as those in Arnol village, demonstrate, through the presence of oars, spars, planks, hatches, farlins, ribs and foreign hardwoods, that driftwood was the predominant source (Walker and McGregor 1996, 5). In 1883, Donald Martin of Tolsta, Lewis, told the Napier Commission how 'driftwood comes ashore here sometimes which is roofing' (cited Cameron 1986, 105). Certainly, few Outer Hebridean communities of the historic period would have had the disposable income to purchase timber, and locally grown sources were, by then, completely inadequate for structural timber. Further examples, such as Borve, Berneray (Mechan and Walker 1989), and Sorisdale, Coll (Figs 7 and 8), indicate that driftwood was the common source until relatively recently. This may explain why it was the tenants that owned the roof timbers in the Outer Hebrides, and not the landlord, as in most of the rest of Scotland (Buchanan 1793 cited Fenton and Walker 1981, 48). An illustration from the island of Canna shows the continuation of the process of driftwood procurement into the 20th century (Fig. 9), and it is possible that particular beaches or bays would have been areas of traditionally high results, as at Mol Mor, Isle of Scarp, near Harris (Fig. 10). Indeed, oral testimony from Heisker suggests that timber (presumably driftwood) 'seemed to be a lot more plentiful...than it is now' (Lachlan Morrison cited Leitch 1990, 9). One further example, from a famous pair of 18th-century tourists, emphasises the scarcity of timber; after discovering that his wooden cane was missing, Dr Samuel Johnson remarked to his companion, James Boswell, as if to forgive the crime: 'Consider, sir, the value of such a piece of timber here!' (Johnson 1984 (1774), 356).
Figure 8: Detail of roof structure at Sorisdale, Coll. Photo by author
Figure 9: Salvaging timber after storms in Canna, 1947 (Campbell 1984,
monochrome illus. 18)
Fig. 10: Mol Mor, Scarpa, Harris. A traditionally plentiful source of
timber (Duncan 1995, colour plate 8)
Given this background where driftwood was of primary importance, there are exceptions where certain areas benefited from alternative local circumstances. Evidence from the Northern Isles suggests that driftwood may have been combined with imported timber from Norway perhaps as late as the 19th century (Hibbert 1822, 414; Thorteinsson 1976, 19; Fenton, Palsson and Cain 1981). On the Inner Hebridean island of Tiree, for example, oral tradition and estate records show that the locals were permitted to take timber from the Duke of Argyll's woodlands beside Loch Suinart on the mainland into the 19th century (Souness 1985, 84). In a further reference, when visiting Barra in 1816, John Macculloch noted that the boats were 'very peculiar' and were built on the island, from timber purchased from northern traders (1819 cited Branigan and Foster 2002, 135).
Thatching material in the historic period came from various sources such as eel grass, seaweed, straw, rush, marram grass, heather, bracken, broom, iris and even potato shaws (Walker et al. 1996, 57). Choice depended upon local supply, legislation and the wealth of the builder. Barley or oat straw, grown on the croft, was a common thatching material in the Northern Hebrides, but, further south and in Tiree, marram grass or bent was favoured, since it was more hardy (Souness 1985, 87; Maxwell 1996, 33). Cutting marram grass was a hard job and it would take one man five to ten days to cut the 50 to 100 bundles necessary for a normal-sized roof. The collection of materials normally took place between September and March, and the thatch was replaced approximately every two years in Tiree (Souness 1985, 89), although it was more, or sometimes less, regular elsewhere. Heather thatch was very common in some parts and, where scarce, would sometimes be used only for the hipped ends of the house. If long strands were supplied in quantity, then the need for a layer of turfs was negated (Souness 1991, 14).
The gathering of thatching material was sometimes prevented by landlords because of the negative impact on grazing (Cameron 1986, 107). This meant that roofs became leaky, resulting in some of the derogatory comments of Victorian travellers who misunderstood the reasons for the condition of some buildings. Thatching was traditionally a communal activity, often involving a team of six men (Walker and McGregor 1996, 11). Oral history from the village of Arnol tells of the importance of having a watcher to check for incoming squalls and bad weather, since it was important to protect the unfinished thatch from high winds. Each house would have landmarks which, if obscured by the incoming weather, would mark the point where the thatch should be rapidly secured (ibid, 11) Rope for securing the thatch was made from twisted heather or straw (Figs 1114), although this was largely replaced by coir rope, from coconut yarn, by the 20th century. Specific local patterns were used for the roping of thatch (Fig. 15), and also in the use of rope to support the thatch, when timber for sarking was rare (Fig. 16). The use of rope as sarking is important, since it indicates that large roofs need not also require large quantities of timber.
Figure 11: Horsehair rope-making, Lochboisdale, South Uist, 1936 (Russell
2002, 84). Photograph by Werner Kissling
Figure 12: Hay rope-making using a simple twister (corthsagan), South
Boisdale, South Uist, 1936 (Russell 2002, 86). Photograph by Werner Kissling
Figure 13: Rope-making with bent/marram grass, Peninerine, South Uist, 1953
(Russell 2002, 105). Photograph by Werner Kissling
Figure 14: Heather rope-making, South Lochboisdale, South Uist, 1947
(Russell 2002, 92). Photograph by Werner Kissling
Figure 15: Roping patterns from the roof at 42 Arnol, Lewis (Walker et al. 1996, 45)
Figure 16: This illustration shows two different methods of supporting turf
and thatch. Left Parallel heather ropes bound to base purlins (Sidinish,
North Uist). Right parallel ropes passed over purlin and round the back of
short lengths of timber between wall-head and the base purlin (Berneray,
Harris) (Walker et al. 1996, 22).
The collection of organic materials appears to have been split between the sexes, with women providing much of the labour, a situation further reflected in the daily agricultural cycle. Traditionally, however, the male head of a household would build the house and gather the stones. In 1883, John Mathieson from Skye described how it took him six or seven weeks to build his house, with occasional help from his neighbours (cited Cameron 1986, 104). This mixture of community and family labour meant there was not much work for masons or other specialists (Cameron 1986, 99).
© Internet Archaeology/Author(s)
University of York legal statements | Terms and Conditions
| File last updated: Tues Feb 28 2006