The roundhouse dominates the Bronze Age landscape. Building traditions include stone, turf, or a combination of these materials (ScARF Chalcolithic and Bronze Age section 3.3) often with an internal postring. The growing body of evidence means that it is now possible to start assessing building traditions and chronological differences looking at topography and timber resources.
The scarcity of Late Neolithic evidence for houses and the absence of Chalcolithic houses hinders analysis of how and when the roundhouse tradition took over in the Highlands. Our earliest evidence for securely dated roundhouses in the Highlands is from around 1800 BC at Lairg (McCullagh and Tipping 1998). Within the well-dated settlements, artefactual assemblages consist of pottery, coarse stone tools and occasionally bone artefacts. No other Highland Early Bronze Age roundhouses have been fully excavated, although there are Early Bronze Age pits and other occupational remains from a number of other sites including rock shelters (Table 6.3). This is despite funerary and artefactual evidence showing Chalcolithic and Bronze Age activity at an early date (Chapter 6.6). There is also the key question of evidence for when timber, stone or turf (and combination) building traditions came into use; a reliance on organic building materials would leave little trace.
In the past, various typologies have been proposed. Three main regional settlement types have been identified for Scotland: unenclosed platform settlements, ring-ditch settlements and ring-banks (ScARF Chalcolithic and Bronze Age section 2.2; Pope 2015). However, it is unlikely that it is possible to find site-specific or even locally typical building traditions given the size of the Highlands with its diverse landscapes. The issue of rebuilding, which is known to occur on some sites, will also blur the picture. Ring-ditch houses which are generally considered Iron Age also occur in the Late Bronze Age (Bradley 2011, 172), with Highland examples from Home Farm, Portree (MHG51648; Suddaby 2013), and Lairg, for example House 2, dated to c 1500 BC (McCullagh and Tipping 1998). There are different formations of ring ditches, however, and difficulties occur in interpretation when these formations are lumped together in a single term.
Altogether there is therefore a need for further well-excavated, dated Highland settlement sites in order to see if there are differences over the larger Highland area. These investigations should also focus on available building materials which are likely to have been the major factor behind building styles. An area without woodland is more likely to use turf; if wood is used, this provides valuable evidence of the ability to source a scarce resource. Increasing attention has been paid in recent years to identifying turf construction (Romankiewicz 2019), and more attention should be focussed on Bronze Age sites.
Some areas, particularly in the upland, preserve large numbers of roundhouses, whereas –unsurprisingly – relatively few have been recorded in the improved agricultural lowlands. One exception can be found at Cnoc Stanger, where excavations showed construction and cultivation on machair (Mercer 1996). Additionally, the WeDigs and Strath Suardal community projects showed that round houses have a longer chronological span in the Highlands than was sometimes thought, with some possibly dating to the Neolithic and others used into the medieval period (Wildgoose 2016; Welti and Wildgoose nd). As a result, area excavation with dating is essential.
Interpretation of the roundhouses as homes is generally based on the presence of a hearth and artefactual evidence, but a number of excavated roundhouses in the Highlands have few if any finds (Cavers et al 2016, 23). The evidence suggests that not all roundhouses were homes.
In general, the buildings from Bronze Age Scotland, including the Highlands, seem designed to provide an internal space of about 8 to 10m in diameter. The ground plan is generally round or slightly oval. The entrance is usually on the south or southeastern side. Roundhouses usually occur as single entities though some conjoined groups are known. They are usually found within dispersed clusters and on the lower slopes, seldom higher than 400m above sea level and seldom on the flood plains of straths and glens. Yet within each of these traits close scrutiny will produce seemingly endless variation (Harris 1985; Mercer 1985; RCAHMS 1993; O’Sullivan 1998; Pope 2015).
At Lairg, detailed survey and excavation showed that there was a considerable degree of variability in size, construction and design within clusters of buildings. This variability was not easily explained by any apparent difference in date or function, though a few structures were monumental with deep entrance passages and were pedestalled by contemporary and/or later tillage. Some buildings were surrounded by an earth and stone bank that often represented multiple phases of build or accumulation. Others were mostly wooden walled except for some stonework near the entrance. Size varied considerably, and House 8 – a circle of small postholes 5m across – represents the most ephemeral architecture of the group. This small building was put up on ploughed land near to the most monumental of the round buildings, but excavation failed to determine whether it had a storage function or served as accommodation for seasonal or slave workers, or for some other socially excluded group. Several of the buildings had large quartz boulders placed to one side of the entrance. An upturned saddle quern against the internal wall directly opposite the entrance may have represented closure and abandonment. The Lairg excavation focussed on chronology and not on architectural form or function. Information relevant to the latter questions will be available on most sites but require specialist skills and investigation techniques to produce reliable observations.
Without doubt, the construction and maintenance of Bronze Age buildings must have been costly to the builders and to their environment, yet archaeologists have not detected the adverse impacts of this architecture. The consequences of thatching such large roofs may well have been greater than the cost of fuelling and feeding the inhabitants. We neither know whether the buildings that we see in survey were for all or just for a fortunate few. Nor do we know whether we are detecting the full spectrum of architecture.
The form of the key features in the buildings varies subtly from building to building, including in features such as the shape and construction of the hearth or doorway. The locations and placements of artefacts are also of importance and perhaps should be considered as artefacts in their own right, for example as part of an abandonment ritual. But perhaps greater significance ought to be applied to the locations and associations of artefacts in buildings that have burned down in use. If the building can be demonstrated to have been abandoned without forethought, as in a fire, then detailed recording of stone tool and pottery assemblages, especially of adjoining fragments, is certainly useful. Unfortunately, at present, archaeologists have few Bronze Age buildings in the Highlands which show evidence of burning down.
The investigations of roundhouses in the Highlands have primarily occurred at Lairg, in the west, and in areas which are today’s marginal land. Although fewer are known from the eastern Highlands, especially on the lowland sites, information is starting to appear from developer-led excavations on the Black Isle and Inverness. In particular, the evidence from Culduthel, Inverness (MHG56078; Hatherley forthcoming), and across the firth at Bellfield, North Kessock (MHG58023; MHG53531; Hatherley and Scholma-Mason forthcoming), should provide useful data, including on whether roundhouses within the settlement were contemporary.
Few Middle Bronze Age buildings have been excavated in the Highlands. A Middle-Late Bronze Age unenclosed platform settlement at Kilearnan Hill, Sutherland, had evidence for post-built structures; this site was part of a long-lived landscape in which burnt mounds and cairns were also constructed (MHG9986; McIntyre 1998). Ring-bank structures are known from both upland and lowland areas, including Lairg, Navidale, Cnoc Stanger and Upper Suisgill. It is noticeable, however, that much of this evidence is confined to the eastern Highlands. Some, such as Cnoc Stanger in Caithness and Upper Suisgill, Sutherland, show repeated rebuilding which again may represent episodic occupation (ScARF Chalcolithic and Bronze Age section 2.2). At Navidale, however, the evidence points to a fairly short span of occupation, c 1400–1200 BC (Dunbar 2007, 159). The roundhouse excavated at Rhiconich, northwest Sutherland, provides useful information about construction and environment in this part of the Highlands which is so often lacking in other data (MHG12143; Donnelly et al 1997; Case Study Rhiconich Roundhouse).
Post-built structures are the main feature of lowland landscapes in the Late Bronze Age. After 1000 BC settlement became even more coastal (ScARF Chalcolithic and Bronze Age panel 2.2). While the Bronze Age settlement at Lairg seems to end or have moved elsewhere during this period, several Late Bronze Age houses were identified on Skye, and in Wester Ross, Upper Suisgill and Connigall in Sutherland, Inverness, and Nairn (Table 6.3).
There are hints of possible non-round construction in the Highlands. Excavations at Garbeg, Inverness-shire, investigated three buildings clustered near the Pictish cemetery where they hoped to find structures associated with the burials. These building were subrectangular, not round, with stone and turf walls (Noble and Sveinbarnarson 2015, 117). The radiocarbon samples surprisingly yielded Bronze Age dates, but until the report is published, it remains an open question as to whether the material tested is either redeposited turf or part of a different style of Bronze Age building in this area. At Connagill, Sutherland, a subrectangular cell built against a roundhouse wall held large concentrations of pottery and bone, some human, dating to the Middle Bronze Age (Dagg 2014). The interpretation for this building is also awaited in the final report.