4.1 Food and Drink

Introduction

Human beings have highly complex relationships with their food and diet, as a diverse range of factors including personal identity, ethnicity and religion all strongly influence the choices people make about what they eat. Hence the subject is one of prime importance within prehistoric studies, particularly those of the Neolithic, which has long been defined, in part, as marking a shift from a hunting, fishing and gathering economy to the adoption of agriculture. Indeed, the question of how, when and why such economic change happened is the topic of considerable theoretical debate (e.g. Thomas 2003; Rowley-Conwy 2004) and Scotland and other marginal European locations often feature prominently in these dialogues. Archaeological data from excavated sites, and the various palaeodietary proxies from the archaeological sciences, offer an increasingly detailed yet occasionally contradictory view of this complex problem. This contribution reviews these data, aiming to assess current state of knowledge of the subsistence practices, economy and the diet during the Neolithic period in Scotland.

Archaeological evidence of diet and subsistence

Cereals and other plant foods

Direct evidence for the use of specific foods is found in the remains of plant and animal foods recorded during excavations of settlements and post-excavation analyses. Plant foods are highly prone to deterioration and unless their context is desiccated, frozen or waterlogged, plants will only survive if the remains themselves have undergone low temperature charring prior to deposition (Hillman et al. 1993). This situation leads to inherent biases in the archaeological record and an under-representation of plant foods in general, especially as the recovery of plant remains from archaeological sites is dependent on excavation methodology and the wet-sieving of soil samples. Unfortunately for Neolithic studies, the most important types of plant remains—specifically cereals and hazelnut shells—are subject to different taphonomic filters, and their recovery is dependent on different excavation methodologies (Jones 2000). Given that cereals are a strong indicator of arable agriculture, whereas hazelnuts are an indicator of the use of (presumably) wild food resources by Neolithic people, the economic and cultural implications of palaeobotanical data has been the subject of considerable debate in Scotland and elsewhere in northern Europe (e.g. Stevens 2007).  For a full discussion of evidence for the use of cereals, fruits and nuts across Scotland in the Neolithic, see Bishop et al.2009).

In Neolithic Scotland, cereal remains have been identified from, amongst others, Balbridie (Fairweather and Ralston 1993), Boghead (MacLean and Rowley-Conwy 1984), Cowie Road (Holden 1997), Claish (Miller and Ramsay 2002), Easterton and Eday (Boyd 1988), Isbister (Lynch 1983), Kinbeachie (Barclay et al. 2001), Knap of Howar (Dickson 1983), Laigh Newton (Toolis 2011), Lairg (Holden 1998), Scord of Brouster (Milles 1986 – although see above, 3.3.7, on the dating of the Scord of Brouster houses), Skara Brae (Dickson and Dickson 2000), Stones of Stenness (MacLean 1978), Toft’s Ness (Dickson and Dickson 2000), Townhead and Unstan (both Boyd 1988) and Warren Field (Murray et al. 2009), and see also Jones and Rowley-Conwy 2007 for further data and references. A number of recently discovered sites, such as Broomhouses (Kirby 2006) and Station Brae (Addyman et al. 2004), have also produced cereals from their excavations, although full details of these are not available at present. Virtually all of these sites have also produced hazelnut shells and occasional remains of other wild food resources such as the crab apple (Dickson and Dickson 2000). Indeed, relative to later prehistoric and historic-period occupation sites, Neolithic sites tend to produce greater quantities of hazelnut fragments, hence implying that hazelnuts were relatively more important in the Neolithic than in later periods (Stevens 2007). Regional variation has been identified in the reliance on domestic as opposed to wild, gathered resources, with a suggestion that much more use was made of wild resources in southern Scotland than the north and west (Bishop et al. 2009). It is clear, however, that cereal foods were present in Scotland from the earliest phases of the Neolithic, and their use continued throughout prehistory. Indeed, the Scottish data even suggest a pattern in the types of crop grown, with barley dominating samples from sites in the wind-swept and wet North, and wheat becoming more prominent in the sunnier and dryer East of the country (Dickson and Dickson 2000). Evidence for the use of oats is rarely found. However, given the interpretative difficulties associated with archaeobotanical data and cereals in particular, there is little consensus regarding the importance of cereal cultivation within the Neolithic economy (Jones and Rowley-Conwy 2007). Notwithstanding the degree of its importance, it can be said with some certainty that cereal agriculture was at least present throughout the Neolithic period in Scotland, and geographically widespread (Table 3). Furthermore, palaeoenvironmental research by Lancaster et al. at Warren Field, Crathes (2009) has indicated that one or more cereal fields or plots existed in the vicinity of the large ‘hall’ there.

Faunal remains

Bone assemblages allow the archaeologist to identify the various animals exploited (whether it be for food or other purposes) by humans at a given place and time; cut marks and fracture patterns inform on ways in which animals had been used, and mortality patterns can reveal stock-keeping strategies. However, due to the high soil acidity over much of Scotland, bones rarely survive, which means that discussion is restricted to a handful of sites, mainly in Orkney.

Recent discoveries of cattle skulls within wall voids at Links of Noltland (Moore & Wilson 2011) and the tibia of hundreds of cattle in a spread surrounding structure 10 at Ness of Brodgar (Card 2010) suggest cattle had important symbolic significance for Orcadian society, but are unlikely to be typical deposits or relate to everyday diet. McCormick and Buckland (1997) and Dickson and Dickson (2000) have summarised the available evidence, and domestic sheep and cattle tend to dominate domestic bone assemblages in equal number, with domesticated pig, wild red deer, aquatic mammals, birds, fish and molluscs being of secondary importance in Orkney. This is true for settlements (such as Skara Brae, Knap of Howar and Tofts Ness) as well as burial monuments such as Ibister, Holm of Papa Westray and Quanterness although it could be argued that the ritualistic aspect of the latter sites means that normal patterns of subsistence may not be represented here.

Papa Westray cattle are much larger than those typical of the Early Neolithic and may represent the presence of wild aurochsen, although cattle bones from other locations are within the normal size range for domesticated cattle (McCormick and Buckland 1997). The very fragmentary remains from Scord of Brouster, Shetland seem to conform to the pattern of sheep and cattle seen in Orcadian sites (ibid. although see also Section 3.3.7 on the dating of the remains from this site). At Northton, on Harris, the Minimum Number of Individual estimates for the site’s small faunal assemblage suggested that sheep predominated over cattle; other species present comprised pig, red deer, seal, whale, fish and seabirds (Simpson et al. 2006). In short, it can be said with some confidence that domesticated livestock-especially cattle and sheep-were used throughout the Neolithic period in Scotland. Red deer and aquatic mammals were exploited too, although it is not clear if this was primarily for their meat, or other products such as their hides, antler or blubber, of which the latter would have been useful as a source of oil for lighting, especially during the months of winter.

As discussed below, scientific analyses of human bones and teeth are invaluable indicators of aspects of diet, and these techniques can also be applied to faunal remains. Balasse et al. (2006) have performed a stable isotope study on sheep teeth from two Orkney sites (Holm of Papa Westray North and the Knap of Howar). Samples from Holm of Papa Westray indicate a partial reliance on seaweed as fodder at certain times of the year  (Balasse & Tresset 2009).

Fish bone assemblages are known from Knap of Howar, Skara Brae and Tofts Ness and a number of burial sites in Orkney,  (for instance Point of Cott and Holm of Papa Westray North) but these pose interpretative problems, partly because their recovery depends on excavation procedure, and their presence at sites (especially burial sites) may be the result of the activities of otters or birds, not humans (Barrett et al. 1999). In contrast to Mesolithic sites in the archipelago, Neolithic fishing activity seems to have been low-intensity, occurring throughout the year (Parks 2009).

Artefacts

The presence or absence of certain technologies, such as pottery, has implications for the cooking and food-storage capacity of past societies. Pottery in particular has considerable potential for the direct detection of food items in the form of impressions, residues and lipid analysis (e.g. Copley et al. 2005; Evans and Bick 1976). Cereal grain impressions are sometimes found on Neolithic pottery-an observation that implies pottery manufacture and cereal agriculture were coincident activities, although few examples of these have been securely dated (Gibson 2002; Brown 2007). Lipid analysis aims to identify certain fatty residues that become absorbed into the fabric of pottery as it is actually used. Lipids tend to become so well-preserved within pottery that their extraction and identification is possible using gas chromatography and mass spectrometry. The analysis that has already been undertaken (by researchers working with Professor Richard Evershed of Bristol University) has shown that dairying seems to have been a component of Neolithic subsistence from the outset, as in Southern Britain. This implies the use of secondary products from cattle, a development that hints at sophisticated stock-keeping strategies practised by people from the Early Neolithic onwards (cf. Sherratt 1981). It also implies that the human population itself may have been either biologically or technologically adapted to this diet, as cows’ milk is not easily digestible by human adults unless they are tolerant to lactose or the milk is further processed and refined.

Other artefacts provide information about the lifeways of Neolithic people; a number of flint tools have functional aspects related to subsistence, for example. Flaked stone bars are likely to have been used during the Neolithic, as later, as ard shares, while the use of spades and hoes is suggested by the marks they have left in sandy soils (c.f. Section 4.1.2.4). Querns, which are a key part of the tool-kit used to process cereals into digestible food, are somewhat under-discussed, perhaps because significant dating problems are presented by their simplicity of form, bulk and their tendency to appear out of context. Massive Neolithic saddle querns are known from Knap of Howar and Skara Brae on Orkney and from Eilean Domhnuill on North Uist where their association with domestic life is clear. An assemblage of quernstone fragments, recovered from pits associated with the Cowie settlement, Stirling, included saddle, trough and saucer querns (Atkinson 2002); these could be interpreted as rubbish, or structured deposits. As for the rubbing stones that would have been used with these querns, both slab rubbers and ball-shaped rubbers are known (e.g. at Eilean Domhnuill). Other coarse stone tools relating to food processing include some unusual tools found at Knap of Howar, which could have been used to pound a specific foodstuff (Ritchie 1983).

Image to Follow

Archaeological features and deposits

At North Mains in Perth and Kinross, cultivation ridges pre-dating a Late Neolithic/Early Bronze Age barrow were discovered by Barclay (1983). A study of soil chemistry and micromorphology suggested that an episode of cultivation, which probably involved intensive soil management techniques such as manuring, also pre-dated a subsequent episode of stock-grazing at the site (Romans and Robertson 1983). The cultivation ridges were placed 2m apart, comparable to “cord-rig” systems more widely known from Bronze Age and later contexts. Because so few surviving deposits of Neolithic palaeosoil have been analysed micromorphologically, it is difficult to establish whether the North Mains cultivation practices were more widespread. Future work could clearly benefit research of this kind; buried soils of Neolithic date are sometimes found on sites like North Mains, or beneath the banks of henges and they have often been used as termini post quos when dating a site’s construction (Davidson and Carter 1997). Such buried soils may have been in part midden material. Although middens have been identified around, even piled against, Orcadian houses, midden deposits have yet to be securely identified in mainland contexts. The recent identification of Neolithic middens in southern Engand in the plough-zone (Cotton & Field 2004) suggests these have been, or could be, found in Scotland.

Although quite rare, a few examples of Neolithic field systems are known; at Scord of Brouster in Shetland (Whittle et al. 1986) and a small number of sites on Arran (Barber 1997), although the dating of the latter is less secure and, as discussed in Section 3.3.7, much of the evidence that had previously been assumed to be of Neolithic date actually belongs to the Bronze Age. At Scord of Brouster, intact field systems were discovered preserved beneath peat, together with evidence of domestic dwellings and macroscopic remains of domesticated plants and animals. Similarly, the Arran field systems are associated with clearance cairns, small buildings and linear marks suggesting ard tillage. Ard and spade marks are also present at the Links of Noltand in the Late Neolithic period, where a ditch rather than a wall was used as a boundary (Clarke and Sharples 1985), and undated ard marks were found underlying the Bronze Age ceremonial monument at Halmie, Highland (Morrison and Pollard 1994). A possible Early Neolithic boundary ditch is also known from Shurton Hill in Shetland (Whittington 1978). Collaborative proxy evidence of land clearance events exists in the form of pollen records (discussed elsewhere) and lake sediments, which reveal discrete episodes of soil erosion during the Neolithic, probably caused by land clearance and tillage for early agriculture (Edwards and Whittington 2001).

Outstanding research questions:

1. How much can be said about the diet (and lifestyle) of people in the Neolithic? The synthesis of environmental information, pollen and excavated material is invaluable, but the development of lipid, isotopic and other forms of analysis that can shed light on diet, mobility, cooking and consumption will potentially offer an invaluable tool. What people ate is of course important (and has some bearing on the transition to the Neolithic) but the variability of food consumption based on age and gender may well become more apparent through further analysis of bone assemblages, pottery and even lithics from past excavations.

See also the Conclusions for section 4.2. Human Remains for discussion of further research questions for food and drink.