One of the enduring questions we have about the Neolithic people who inhabited Scotland is: where were they living? On the face of it, this seems like a simple question, but underlying it is a complex range of evidence which suggests this question cannot easily be considered without recourse to thinking about wider economic and subsistence patterns. Any research framework for Scotland's Neolithic has to answer this question, but also consider related issues (often downplayed in Neolithic studies in Britain and Ireland): What was it like to live in the Neolithic? What happened on a daily basis? Can the 'domestic' be identified in the archaeological record, and what might it look like? In this section the range of evidence available for domestic structures and everyday life in the Neolithic in Scotland is reviewed (for more detailed summaries and site descriptions see Barclay 1996, 2003 and Brophy 2006, forthcoming).
The problem with Neolithic houses
There is no doubt that people in the Neolithic of Scotland were living somewhere, but exactly where, and in what type of structure, has become a subject of some debate within Neolithic studies for Britain and Ireland as a whole (cf. Darvill and Thomas 1996). From the outset of Neolithic studies, there was an expectation that Neolithic settlement in the British Isles would look rather like the timber longhouse settlements that defined the early Neolithic Linearbandkeramik of central Europe. A lack of any apparent traditions of timber houses in much of Britain gradually eroded this confidence. Nonetheless, the expectation of a Neolithic of farmhouses, fields and sedentism continued to prevail until recent decades (Bradley 2003; Gibson 2003). The lack of buildings discovered in the archaeological record was explained in a number of ways (Darvill 1996). For instance, Kinnes (1985) suggested that virtually no Neolithic buildings had been found in mainland Scotland because the nature of these buildings may have left little or no traces in the archaeological record, and more recently the use of turf as a building material has been postulated (Loveday 2006).
Despite the excavation of the unusual 'village' at Skara Brae, Orkney, little progress was made in identifying wider patterns of settlement, in particular mainland Scotland (Barclay 2003). The excavation of a cropmark site at Balbridie, Aberdeenshire in the 1970s (Fairweather and Ralston 1993) offered some comfort for those looking for Continental-style longhouses in Scotland, although the nature and function of this and similar buildings is still disputed. However, with an absence of many other comparable structures in mainland Britain as a whole, and within the context of a new ideological model of the Neolithic developed in the 1980s, the absence of evidence gradually began to be taken as substantial evidence for absence. Subsistence models were developed that suggested that (early) Neolithic farmers were at least semi- (if not wholly) mobile, moving between temporary camps and monuments. This position was not viewed as being mutually exclusive with low level cereal farming and animal herding. This position became known as the 'mobile Neolithic hypothesis' and became something of an orthodoxy among some people by the mid 1990s(e.g. Thomas 1996; Gibson 2003 and see Barclay 2001). Within this context, the use of 'loaded' terms like 'farm', 'house' and 'domestic' were avoided, and replaced with more neutral terms such as building and structure (Brophy forthcoming).
More recently, evidence from developer-funded excavations (and to a lesser extent aerial reconnaissance) in both Ireland and Scotland has lead to arguments that a mobile Neolithic model may not be entirely appropriate for all of the British Isles (cf. Cooney 2003; Barclay 2001; Cross 2003; Smyth 2006; 2012). As we shall see, in Scotland a wide range of putative settlement sites and buildings have now been identified, which suggest different degrees of sedentism may have been the norm in Neolithic Scotland, with clear regional variation (Brophy 2006). There is still a realisation that expecting to find wholly domestic activity may be inappropriate for Neolithic contexts, but also recognition that ritualised aspects of buildings, pit digging and deposition do not preclude their association with everyday life and daily routine. For the remainder of this section, then, evidence for a range of possibly domestic sites will be considered from across Scotland, from the unique to the typical.
Island living, stone houses
Scotland is fortunate to have world famous and incredibly well preserved houses built from stone. Unfortunately, these have a reasonably limited distribution, being found almost entirely in the Northern Isles. The study of stone houses and 'villages' contrasts sharply between Orkney and Shetland; while the former has been the focus of various high profile excavations, the resource in Shetland remains largely 'untapped'. There are also a series of important and distinctive Neolithic settlement sites in the Western Isles.
A number of Neolithic settlements have been excavated in Orkney. These are largely dated to the later Neolithic, although early Neolithic precursors to later house forms include Knap of Howar, Papa Westray (Ritchie 1983) and Wideford, Mainland (unpublished). The later Neolithic 'village' site of Skara Brae (Childe 1931; Clarke 1976, 2003) is well known, but to this must be added an impressive list of other settlement sites that have been excavated on mainland Orkney - Barnhouse (Richards 2005), Crossiecrown (see Jones et al. 2010) and Ness of Brodgar (the latter two as yet unpublished). Other island (non Mainland) settlements excavated include Pool, Sanday (Hunter 2000), Links of Noltland, Westray (cf. Clarke and Sharples 1985; Moore and Wilson 2011), Rinyo, Rousay (Childe and Grant 1939), Ha’Breck, Wyre and Green, Eday (latter two unpublished). These clusters of buildings represent long-lived multi-phase settlements, and the stone construction of the buildings has facilitated a general understanding of the layout of the later Neolithic house in Orkney. The small buildings were arranged around a central hearth, with evidence for beds and storage space (e.g. 'dressers' and cupboard space) in some buildings. A series of atypical or specialist buildings – workshops, double houses, 'shrines', 'temples' and large communal buildings – has also been identified at various sites. The houses within these 'villages' have different arrangements; at Skara Brae the houses were closely clustered and connected by low passages (semi-subterranean, surrounded by midden), while at Barnhouse the houses were slightly more dispersed and arranged around a central yard. A series of ancillary features have also been identified, such as elaborate drainage systems, thick walls and outdoor hearths. Settlements were also characterised by long-term use, repairs and replacement buildings. The recent identification of timber buildings, underlying stone phases of construction, at Wideford Hill (round structures) and Ha’Breck (rectangular) suggests the first farmers on Orkney may have inhabited timber houses similar to those used on the mainland (see below).
The survival of these buildings, either preserved beneath sand (Skara Brae) or in the ploughsoil (Barnhouse, Ness of Brodgar), has allowed some insight into their architecture. The buildings were typically constructed of local flags, with orthostats and slabs defining furniture. Buildings at some sites are embedded within midden material, giving a semi-subterranean effect. There is no indication of roof form or material, but the houses are presumed to have had pitched roofs, perhaps with timber supports and lined with turves, straw or seaweed. The nature of the buildings has generally been accepted as domestic; typical buildings would have housed a small family group. Richards (2005) has argued that these domestic spaces embodied a shared cosmology within their architecture that can also be recognised at the contemporary ceremonial and burial monuments of Orkney. This architectural template seems to have reached its peak with the massively enlarged house form of the structure 10 ‘temple’ at Ness of Brodgar (Card 2010).
As with evidence for diet and subsistence, a disproportionate amount of current information about Neolithic settlement in Scotland is based on Orkney and it may well be that this evidence is atypical. The potential for more buildings and settlements is clear, (as attested to by ongoing annual excavations of Neolithic settlements in Orkney) but perhaps a still greater untapped potential exists in Shetland. Calder (1950, 1956) initially highlighted a remarkable quantity of possible Neolithic stone buildings and associated field-systems across Shetland, obscured by peat and later settlement and farming - although, as seen in Section 3.3.7, many of these are more likely to be of post-2500 BC date. Calder identified 57 possible buildings, with a range of architectural forms. He also identified what he argued were related dyke / stone field boundary systems and clearance cairns, and he carried out basic excavations at a few sites. More recently, the Neolithic settlement at Scord of Brouster was excavated (Whittle et al. 1986); once again, however, a critical reassessment of the evidence indicates that most of the evidence there is not of Neolithic date (see 3.3.7)
Although there is some variation in these structures, most are oval in plan (or 'heel-shaped') with thick stone walls. There is little evidence for internal 'furniture' although some have central hearths and drains, while others have recesses arranged around the walls. Timber posts may have supported the roofs, with internal postholes found in some buildings. The most comprehensively excavated settlement, Scord of Brouster, consisted of three buildings, not all in use at the same time. Interestingly, not all buildings were of stone; Whittle et al. (1986, 133) identified phases of timber building, which he characterised as 'flimsy transient shelters'. There is still not a full understanding of the range of buildings identified, typified by enigmatic stone structures such as the Stanydale 'temple' (Calder 1950, Barclay 1996, 65–6), but such sites suggest that as with Orkney a range of specialist and perhaps ceremonial structures were constructed along the same lines as 'domestic' dwellings. For instance, one of the buildings at Scord of Brouster was interpreted as a pen or stock enclosure.
There is equal potential for settlement evidence in the Western Isles, in various forms, but often obscured (and preserved) both by machair and extensive later prehistoric and Norse settlement in the same locations (Armit 2003). Excavations have revealed various settlements of Neolithic date. These include the remarkable site of Eilean Domhnuill, North Uist. This small islet was occupied in the period 3650–2600 cal BC (ibid, 93) through a series of small rectangular buildings with stone foundations, probably turf walls, and central hearth, with associated animal pens, yards and surrounding palisade. This island could only have been occupied seasonally. The untapped potential of such islets / crannogs was further indicated by the discovery of a range of Neolithic material at Loch a’Choire, South Unist (Henley 2012; and Robert Lenfert pers comm). On Harris, Neolithic / beaker structures were excavated at Northton in 1965-66 (Simpson and Murphy 2003; Simpson et al. 2006); the house forms are similar to Eilean Domhnuill. A key characteristics of the Neolithic settlement sites known in the Western Isles is that they usually form only part of very long (several millennia) sequences of occupation, with Neolithic 'phases' at sites such as Alt Chrysal and Bharpa Caranish (ibid).
Most Neolithic buildings in the British Isles that could have had a domestic role were made of timber (cf. Darvill 1996; Smyth 2006; Brophy forthcoming). However, there is considerable variation in size, construction style and function, even within Scotland.
At the larger end of the scale, a small group of rectangular timber halls have been identified as cropmarks by aerial reconnaissance since 1976. There are potentially as many as ten such sites known in the cropmark record, and four have now been excavated: Balbridie, Aberdeenshire (Fairweather and Ralston 1993), Claish, Stirlingshire (Barclay et al. 2002), Warren Field (Murray et al. 2009) and Lockerbie Academy, Dumfries and Galloway (Kirby 2011). Doon Hill A, East Lothian, may also have been a Neolithic hall (Sheridan forthcoming). Unexcavated examples include Noranbank, Angus and Sprouston, Scottish Borders (Smith 1991). These buildings share remarkable similarities, but it is by no means certain that they fulfilled a wholly domestic role. Balbridie for instance has been characterised as anything from a European-style longhouse to a cult house or feasting hall. All were in use from very early in the Neolithic, c3900-3700 cal BC, and were all burnt down (Brophy 2007; Whittle et al. 2011).
The excavations of these sites have given a good impression of their scale, and the use of space within these buildings. The ground plan is typically in the order of up to 25m in length and 12m width, usually with slightly rounded ends. There is no evidence for roof height or form, but there is a suggestion that the roof could have had a maximum height of 8m at Claish and Balbridie and had been pitched (Barclay et al. 2002). The buildings were constructed from large oak posts; in the case of Balbridie these were set into a foundation trench. These may have been supplemented by wattle and daub outer layer, or turf (Loveday 2006), but there is no strong evidence for this. At Claish, there was a suggestion that the walls were repaired, with a few instances of post replacement. Roof support would probably have fallen on internal posts rather than the side walls, and internal divisions and large axial posts were found at all excavated sites. Some of these divisions may have been in the form of continuous panels or screens. Entrance to the buildings was typically through entrance gaps on one or both short end walls, although these entrances were often narrow, perhaps even awkward. An open 'yard' was found at one end of Warren Field and Lockerbie Academy.
Little evidence was found within these buildings for what was going on inside them. Each was associated with Carinated Bowl styles of pottery, although this in itself does not indicate domestic activity. Hearths, or at least burning pits, were found within all buildings, but floor surfaces were not recovered. Internal divisions suggest that space was ordered within each building, perhaps with 'rooms'. It could also be imagined that the roof space was utilised for storage, and perhaps an upper 'floor' was created. It is certainly possible to imagine an extended family (and perhaps animals) living in such a building, although it could also have been used for temporary accommodation (Brophy 2007) or even feasting (hosting up to 50 people (Ashmore 1996)).
However these remarkable buildings are viewed (and there are certainly a few other potential candidates in the cropmark record in eastern Scotland such as Noranbank, Angus and Sprouston, Borders) they seem to have been unusual, not typical, early Neolithic structures. Cross (2003) has argued that these ‘halls’ may have acted as a kind of Scottish equivalent of causewayed enclosures, permanent structures within a mobile and dispersed early Neolithic population. As spectacular buildings, perhaps in wide woodland clearances (Lancaster et al. 2009), they would have been an enduring and imposing permanent presence in the landscape for one or two generations of early farmers. Cross (2003) has argued that these 'halls' may have acted as a kind of equivalent of causewayed enclosures, permanent structures within a mobile and dispersed early Neolithic population.
Camps and timber-framed buildings
If timber halls were unusual structures and belonging to the very beginning of the 'Carinated Bowl Neolithic', where were the majority of people living? There is certainly evidence for smaller scale, less permanent and imposing structures, across mainland Scotland, albeit mostly confined to the 4th millennium BC. In reviews of the evidence, Barclay (cf 1996, 2003) has listed a small number of light timber structures, mostly found during developer-funded excavations or through test-pitting lithic scatters. Although again there is little inherently 'domestic' about these structures, they seem the best candidates for where Neolithic people were living, eating and sleeping. When found, these buildings often take the form of a collection of post-holes, stake-holes, pits and possibly hearths, and sometimes an intuitive 'join-the-dots' approach is required to make sense of this apparent disorder.
In some cases, these could be interpreted as no more than campsites (overnight or periodically used). Examples of these have already been discussed in theme 3 (e.g. at Auchategan, Argyll & Bute). Slightly more sturdy structures have been found although in sum less than twenty Neolithic ‘houses’ are known. The majority seem to be rectangular in plan, with light timber frames, such as Kinbeachie Farm, Highland (Barclay et al. 2001) and Biggar Common, South Lanarkshire (Johnston 1997). The rather more substantial early Neolithic house at Laigh Newton, East Ayrshire, indicates the hidden potential in the plough-zone (Toolis 2011). A number of undated rectangular structures have also been tentatively interpreted as Neolithic houses, such as Kingarth, Bute (Mudie and Richardson 2006) and Ratho Quarry, Edinburgh (Smith 1995). In themselves, this small group of negative features does not amount to a tradition, but suggests that there were light timber-framed buildings in Scotland in the Neolithic. Some are associated with central hearths, and most are in a range of 5m to 10m in length. There is little idea of the permanence or otherwise of these structures, but they conceivably could have held a small family group for a number of years, or were perhaps returned to seasonally.
Several oval to round buildings have also been found, largely dated to the later Neolithic, notably Beckton Farm, Dumfries and Galloway (Pollard 1997) and Cowie, Stirling (Atkinson 2002). In both cases, these later Neolithic settlements consisted of multiple phases of sub-circular stake-built structures, with palisade slots and central hearths. The double walls at Cowie may be associated with insulation or storage, and as noted above, large pits with quern fragments were found in the vicinity of these buildings. The tendency for later Neolithic buildings to be circular rather than rectangular (cf. Bradley 2007) has recently been reinforced by the re-interpretation of Greenbogs, Aberdeenshire. This structure, with four central posts and a circular outer wall, fits well with a wider tradition of ‘Grooved Ware’ houses found across southern Britain; the substantial internal supports of such buildings make them more liable to be found in the cropmark record (Noble et. al. forthcoming). Such structures can also be recognised at Beckton Farm.
The paucity of evidence for 'everyday' houses and settlements is somewhat alleviated by a larger body of data in the form of pits. Pits in themselves are not indicative of settlement, and indeed can be interpreted as evidence for ritual activity. However, they also sometimes contain material which could be interpreted as domestic refuse, and are repositories of good environmental evidence and a rich selection of material culture. Developer-funded archaeology has allowed a wide range of Neolithic pits to be found across the Scottish mainland, often in isolation. In the period 1985-2005, over 50 Neolithic pits or pit groups were found across Scotland (Brophy 2006, 22; Brophy and Noble 2011). Pits have been found that contain cereals, various types of pottery, polished stone axe fragments, carbonised material, smashed quern stones and so on. Although interpreting pits is difficult, there is little doubt that they offer a picture of a general low level of inhabitation of the wider landscape, sometimes found in isolation, in other cases in large groups. ‘Pit clusters’ such as Grandtully, Perth and Kinross (Simpson & Coles 1990) and Meadowend Farm, Clackannanshire (Timpany et al, forthcoming) may well relate to seasonal and / or repeated occupation of certain locations (cf. Garrow et al. 2005). Pits could be viewed as indicative of a range of behaviour and individual actions in the Neolithic, from the explicitly ceremonial to casual rubbish disposal; it is often difficult to differentiate the two in the archaeological record (Pollard 1997; Brophy & Noble 2011). Nonetheless, given the paucity of buildings with occupation levels, these represent a valuable and – as yet - relatively untapped repository of information on everyday life, and the environment, in the Neolithic.
Everyday life in the Neolithic
There is currently a basic understanding, then, of the kinds of structures that people were living in during the Neolithic. This range of buildings probably reflects different temporal scales of settlement, from permanent houses to short-lived campsites. Aside from the buildings themselves, though, what other information have settlement sites provided about everyday life? Aspects of subsistence have been discussed above (Section 4.1), while other activities that might be imagined were being carried out at settlements e.g. pottery manufacture, lithic knapping and other craft activities are covered elsewhere (theme 5), although again direct evidence for this is limited. It might also be naïve to expect to find all evidence for such activities within houses; it may have been more appropriate for them to happen at arms' length from domestic spaces (Whittle et al. 1986, 136), perhaps in specialist buildings or zones within a settlement. One such structure has been suggested as a locus for knapping at Skara Brae, while the occasional presence of burnt lumps of potter's clay in settlements confirms that potting was carried out in the vicinity of dwelling structures, if not inside them.
Whittle (2003) has considered what he called the 'daily round', evidence for the kinds of activities Neolithic people were carrying out in and around settlements. In one sense, these routines are what are most intriguing for those with an interest in the past, and yet are very difficult to get to through the archaeological record. For instance, outside of Orkney, there is virtually no evidence for a range of bodily requirements, (for food, see Section 4.1). Stone bed spaces have been identified at several Orkney settlements including Skara Brae and Barnhouse. These box-like beds are defined by orthostats and will have held organic bedding of some kind. A further model, based on much more recent analogies from the Scottish Highlands would see such as space as being used for sleeping by a group of individuals, as Childe imagined (1931). The so-called 'dressers' found within some Neolithic houses in Orkney seem to have served a storage or display function; this may though have been more subtle than simply displaying 'best Grooved Ware', and instead acted as a ceremonial focus (or shrine?) for the occupants of the house. An enlarged dresser utilising differently coloured sandstone slabs found within structure 10, Ness of Brodgar, further hints at the key significance of such ‘dressers’ within Orcadian houses as well as the exceptional nature of this building (Card 2010). No such furniture has been identified at any non-stone Neolithic building (or indeed any floor surfaces) in Scotland although the presence of beam slots for box beds and possibly dressers in some houses excavated beneath Durrington Walls henge, Wiltshire suggests that features are not impossible to find in timber structures (Parker Pearson et al. 2007).
Evidence for toilet activities is still rarer. The usual environmental analysis of the fills of pits associated with, or near, settlements may not pick up traces of such activity, but pits may be one element of this aspect of everyday life. While a high quantity of organic material was found within Pit 1, Cowie, this feature was probably Mesolithic (Atkinson 2002); phosphate analysis of pit fills at Cowie has not been repeated at many other similar sites. Childe (1931, 18) argued that recesses in the corners of buildings within Skara Brae may have been used as 'privvies', although this idea had been dismissed by Richards (2005, 123). However, a complex series of drains discovered at Barnhouse led Richards to suggest that the inhabitants of the village had a specific concern with disposing of liquid waste, as well as channelling rain water (2005). Drains were also found at Skara Brae and Rinyo. Subsequent work has developed these ideas. Detailed geochemical and soil micromorphological analysis of floor surfaces at House 1, Crossiecrown, Orkney, suggested a drain in one corner of this building was a place where deposits, including excrement and hearth ash, accumulated (Jones et al. 2010). The survival of floor surfaces in vital here, and the opportunity for relevant analysis is possible at least in Orkney although less likely elsewhere. Ultimately, such activities may have taken place in the woodlands (for which there is good anthropological evidence (Whittle 2003) that surrounded many Neolithic settlements in mainland Scotland.
Phosphate analysis at Crossiecrown, copyright R Jones.
Activities and movement within buildings may be possible to pick up in instances of exceptional preservation (e.g. hollows associated with squatting beside a hearth at Durrington Walls) but as yet no such evidence has been found in Scotland. Certainly, where floor surfaces or evidence survive, houses seem to have been arranged around firespots or hearths. Evidence for in situ burning in hearths in Orkney, as well as the so-called 'fire-pit' within Claish timber hall, suggest that these buildings may well have been smoky places (some structures may have been used as ‘sweat lodges’ (Jim Leary pers comm.). Logic dictates that opening in the roof or 'chimneys' were unlikely, and it is more likely that smoke would have been able to dissipate into roof spaces. Both wood and peat were burned within hearths at Scord of Brouster (Whittle et al. 1986, 133) while there is also a possibility that some stone 'troughs' within Orkney houses may have been associated with heating stones for cooking. Interestingly, the discovery of 'hearths' outwith buildings at Barnhouse and Beckton Farm suggests that some activities based around fires – e.g. cooking, storytelling and crafts as well as pottery firing – may have been communal in nature; the external hearth at Barnhouse for instance was very large and on the edge of the settlement (Richards 2005). Fires within houses may not have been used for cooking or at least served other roles, such as smoking and preserving food stored in the roof space, keeping insects and rodents at bay, ceremonial burning and so on.
The subdivision of space at a range of buildings – from timber halls to the Orkney houses suggests differential use of space within these structures, perhaps rooms. Topping (1997) has argued that such spatial divisions may have had social or ideological meaning, although there may have been more prosaic reasons for dividing up space (e.g. animals in the building, although there is as yet no strong evidence for this in Scotland). In a society where there probably would have been no clear divisions between ritual and domestic, houses may well have been a focus for routine and ritualised actions, non-domestic activities, deposition and ceremonial activities, activities which may have been private. Evidence for such activities has found at a range of Neolithic buildings in Scotland. The deposition of unusual and / or large quantities of grain at both Balbridie timber hall and Barnhouse house 8 and Laigh Newton have been interpreted as either offerings, or indicative of a grain storage role (and both are huge buildings). Pollard (1997) and Atkinson (2002) have both developed ideas of routine and ritual deposition and activity within a settlement context; the two may not have been distinct in the Neolithic. Further boundaries may have been eroded with the placement of human remains in 'domestic' contexts, attested to in various forms at Skara Brae, Beckton and Raigmore.
When everyday life in the Neolithic is considered, there are more questions than answers. The key outstanding research questions can be identified as:
- More better-preserved structures need to be found outside of the Northern Isles, in order to correct the current geographical imbalance in the evidence
- Within Orkney, the early timber buildings need to be understood more fully
- What range of structures were used for habitation at different times during the Neolithic, and how do these relate to overall patterns of land use and subsistence? Can any patterns relating to social differentiation be identified in the domestic arena?
- There is clearly a regional tradition of building with stone in areas where timber was scarce; but can other regional traditions be identified? To what extent was peat used as a building material?
- Can any trends over time be identified in either the shape and layout or construction of dwelling structures?
- Can activity areas and 'taskscapes' be identified with greater clarity?
- Can socially meaningful and ritualised activities be identified within settlements?