3.5 Living with Animals and Sustaining Cereals

A remaining gap in current knowledge is archaeological data on animal husbandry, particularly in rural areas earlier in the period. Notable exceptions are the midden site at Eyemouth in Berwickshire (Dixon 1986), Freswick Links, Caithness (Batey, Morris and Rackham 1995), Bornais in South Uist (Sharples 2005), and the recently excavated Eldbotle village, East Lothian – all of them coastal settlements preserved by sand. The highland and upland areas of Scotland are particularly poor in this respect, largely due to poor bone preservation, areas in which animal husbandry predominated over arable production. Similarly, archaeological data for crop production is limited to the Lowlands and the Outer Isles. Evidence from later urban sites is much more plentiful (e.g. at Perth, Hodgson et al. 2011)

Although zoologists had studied bones from early excavations of the late 19th century, which were reported on in the Proceedings of the Society of Antiquaries of Scotland, the earliest approach was merely to identify the species present, without quantifying the material (eg Smith 1878). In most cases only a non-random sample of bones appears to have been studied – for example, particular parts of the skeleton only.  These early reports are tantalising, particularly when it is appreciated that most of the material described was not considered of sufficient value to warrant storing it for the future.  However, from the 1930s onwards, specialists, foremost among them Marjory Platt, were studying bones from Ministry of Works excavations such as at  Jarlshof and the broch of Midhowe, Rousay and were taking a more analytical approach (Platt 1934a, 1934b, 1956). However, although it is clear that by this date metrical data were collected and analysed, the publications often contained only a brief, and in this respect, still sometimes unsatisfactory summary of the results. It was not until the 1970s when the first large scale rescue excavations were carried out that environmental archaeology in Scotland began to develop adequately.

Regardless of geographical location, early medieval animal assemblages are scanty. Of those which have produced large quantities of bone, the multi-period castles site at Dunbar is probably the most prolific (Perry 2000), while results from Portmahomack are eagerly awaited. Bones have not survived well at any of the Alcocks’ (1992) excavations of early sites (for example Dunottar and Dunollie,) while those from Urquhart Castle are probably from the later medieval or perhaps post-medieval period (Smith unpublished a).

In the Western Isles the large assemblages of bone recovered from the excavations at Bornais and Cille Pheadair indicate the importance of sheep though only to a slightly greater extent than cattle (Sharples et al forthcoming). Surprisingly pig and red deer are a significant feature of the assemblages throughout the period and pigs are particularly important in the Viking period which is a pattern repeated in the Faroes and Iceland and may represent the importance of feasting. The importance of fish in the Viking economy is also demonstrated by a substantial increase in herring fishing (Sharples 2005) which contrasts dramatically with the pre-Norse interest in saithe and salmon and probably indicates a shift from line fishing from the shore to deep sea fishing from boats.

The geographical coverage of the mainland is, therefore, patchy.  The north-east has been more widely studied than the west, and the far north hardly at all, a result of the uneven nature of rescue archaeology, and the variable preservation of environmental remains over Scotland. More acid soils, such as occur in the uplands and in the west, are far less conducive to preservation of bone.

Given these limitations, it has been possible to use environmental data to gain an insight into that part of the medieval economy based on primary products derived from animals and plants. Animal products would include bone, hides, wool, woolfells as well as foodstuffs such as meat, salmon, and salt fish. Plant remains indicate the the prevalence of barley and oats with some wheat as cereals, flax, hemp, pulses, fodder crops and a vast range of herbs, roots and leaves that would serve curative, flavouring as well as nutritional purposes. Many of these  appear in later medieval documentary sources such as the Exchequer Rolls and the Rental Books of religious houses. Results from Perth appear to suggest an 11^th/12^th century economy based mainly on cattle rather than sheep, the bias towards sheep occurring only very gradually with time. It has so far been assumed that this relates both to the high status accorded to the owning of cattle in the early Highlands as well as to the value of hides to the export economy with respect to wool. Further south, for example in the Lothians and Borders, sheep would appear to become more numerous at an earlier stage, and it is suspected that this is connected to the economic influence of the great Abbeys (e.g. Jedburgh Friary assemblage, Dixon et al. 2000). In the west, results from Glasgow so far appear to follow the more Highland pattern of a cattle-dominated economy (Smith unpublished b). In Edinburgh, information, apart from the High Street excavations of the 1970s, was sparse until relatively recently, with assemblages having been recovered from sites in the Cowgate and in the Canongate in advance of the Holyrood Parliament building. On the west coast, assemblages are much fewer, the unpublished MSC excavations in Ayr and Glasgow in the 1980s being the exception rather than the rule. In the Highlands, there are only good results from a small number of sites in Inverness, notably Castle Street (Wordsworth 1982). Aberdeen and Perth were centres for trade in cattle products and Roxburgh too was an entrepot, while Berwick was used for the international trade in sheep products, wool and woolfells. As for non-urban sites, these are far fewer across Scotland. Castle sites, such as Urquhart Castle in Inverness-shire, where bone was well preserved, seem to be the exception rather than the rule, although archive reports on other castle assemblages lurk in the grey literature (for example Stirling Castle (Thoms nd)).

Natural resources included turf and peat, wood products including charcoal, water, bog iron, other building materials, clay, stone and coverings like reed or heather, sea coal and salt. Fishing is both a fluvial and coastal activity but rarely in the medieval period one carried out in fishponds due to the abundance of river fish, especially salmon. A variety of structures including cruives or yairs were used to trap fish in the inter-tidal zones and along rivers (Hale 2005). Evidence for the former is better known, through coastal surveys, than the latter, while deep-sea fishing is mainly studied through the resultant fish-processing waste at sites like Eyemouth or Freswick Links, where the calcareous sand environment preserves the bone assemblage (Dixon 1986; Batey et al. 1995).

Developments in scientific techniques (see also the ScARF Science Report) have also changed the way environmental remains are examined. Previously, the comparative anatomical approach, where archaeological specimens were identified by direct comparison with a reference collection, was entirely dominant. Recent advances in scientific approaches include techniques allowing for better ageing of populations (by for example the thin-sectioning of teeth) to DNA analysis, lipid analysis and stable isotope analysis. All of these techniques however require the availability of specialist knowledge, equipment and lab space and the establishment of comparative datasets still remains an essential first step.  For example, an important dating question which has arisen with regard to sites in the Northern Isles (Ballin Smith 1995), is to distinguish red deer from reindeer antler with the requirement to recognise the likely presence of foreign antler by traditional means confirmation.

Goat Horn Cores from Perth High Street. When first excavated, separating horn cores into males, females and castrates was determined by their morphology (appearance) and partial or difficult specimens could only be described as sheep/goat. In future, techniques such as DNA analysis can be used to determine species and gender and stable isotope analysis may be used to locate the animal’s place of origin © Catherine Smith.

Stable isotope analysis has not yet been used to determine the origin of sheep flocks whose bones have been recovered in the burghs, but it has great potential to unlock the relationships between the consuming burghs and the supplying hinterlands. Did great flocks of sheep indeed walk to the markets of Perth from Highland Perthshire, as is suspected? Did the goats of Argyll end up in Glasgow? Can the documentary sources which record border disputes of burghs such as Perth and Auchterarder regarding the extent of their hinterlands be elucidated by data suggesting who was selling produce and where? Testing the chemical origins of the bones may indeed supply some of the answers. Lipid analysis of fats encrusted on pottery is currently also providing clues to industrial processes such as fish liver oil exploitation as well as food production and dairying (Mulville and Outram 2005).

Cattle Mandible from Perth High Street. Stable isotope analysis may be used to locate the place of origin of cattle, sheep and goats and has great potential to unlock the relationships between the consuming burghs and the supplying hinterlands © Catherine Smith.

As with analysis of the bone assemblages, archaeological research into agricultural practices and the range of crops grown in medieval Scotland has been very limited. There is a tendency amongst historians and archaeologists to model medieval Scottish agriculture using English templates, which generally privilege arable regimes and especially the production of cereals, over mixed or principally pastoral regimes. This is reflected in a bias towards recording and analysis of evidence for arable systems (Dixon 1994; Dodgshon 1994).

Most recent work has focused on non-excavational recording of field systems, palynological analysis of soils and sediments and identification of plant remains from a variety of contexts. it has mainly taken place in the south-eastern Lowlands and the Northern and Western Isles and most of mainland Scotland outside of the south east is poorly represented. Geoarchaeaological research is significantly more advanced and more geographically widespread. Recent work in this respect has been undertaken on processes of soil formation (anthrosols, hortosols or technosols) in a variety of urban, peri-urban and rural contexts, but mainly in the north-eastern and south-eastern Lowlands (e.g. Nairn, St Andrews, Lauder and Eldbotle), Hebrides (Pabbay) and Northern Isles (Papa Stour): the Highlands and western Lowlands, however, still remain under-researched (Davidson and Simpson 1994; Carter 2001; Golding, Davidson and Wilson 2010; Oram forthcoming).

Rig, Stewarton Cottage, Scottish Borders. This aerial photograph shows several phases and types of rig cultivation from the medieval and later periods typical of a Lowland context. Elsewhere in Scotland field systems often display different types of rig due to the different techniques of cultivation © HES

Historical sources, principally but not exclusively from monastic and royal estate records, have yielded considerable data on agricultural practice, especially on processes of clearance and manuring of arable land (Golding, Davidson and Wilson 2010; Oram forthcoming), and have shed some light on both the range of generic crops produced and also on processing methods. For example, rye, which is also known from archaeological contexts, has been revealed as an important cereal type grown; wheat is now understood as being a regionally significant crop in parts of Lothian, the Merse and eastern Galloway; and there is extensive reference to malting of oats as well as of barley. How extensive the growth of such species and the use of such practices was in less well documented regions remains to be tested by archaeological investigation. One important deficiency for the modern scholar in most of the medieval records lies in the Latin terminology which does not reveal the variants of barley (bere) and oats (black, grey and white oats) adapted to the Scottish climate.  Such distinctions can be distinguished from archaeological data (Dixon forthcoming) and do begin to appear in late 15th- and early 16th-century records written in Scots. It remains to be established if this represented a recent development made in response to the impacts of long-term climate change or simply clearer recording of long-established practices. Peas have also been recognised as a significant element in Lowland arable crops, well-attested in historical records and from excavated contexts, their presence raising some questions over traditional thinking on medieval soil conditions (e.g nitrogen levels).

While knowledge of field crops grown is far from complete, understanding of medieval Scottish horticulture is woefully deficient. Soil-formation processes in urban backlands, peri-urban fields and fermtoun kailyards is relatively well understood, but what was grown in those soils remains more a subject of tradition than of scientific research.  Documentary sources illustrate the planting of fruit trees; however other than within monastic herb gardens – and even here modern descriptions are based on analogy with English or continental examples rather than empiric  evidence from Scotland – there is an extremely limited understanding of what medieval gardeners/horticulturalists were growing by way of vegetables or herbs. Amongst other ‘crops’ generally overlooked in archaeological and historical analyses of medieval husbandry practices, hay is perhaps the most significant. Over-dependence on Improvement-era literature has led to archaeological and historical acceptance of claims that hay production was negligible or at low level in most Highland and Hebridean areas before the late 18^th century. This view has, consequently, coloured assumptions about livestock management and the over-wintering of animals (McCormick 1998). Research on grazing has tended to focus on the physical structures of shieling huts at summer pastures and less on the ecology of the surrounding grassland (but see, for example, Tipping 1999, 2004) or on the documentary records of souming levels (Ross 2006), over-grazing, climate events and pathogens which had an impact on carrying capacities and, consequently, on livestock numbers over time (Oram and Adderley 2008, 2010a, 2010b). Awareness of the impact of historic climate change on pasture and on fodder supplies is essential for better understanding of shifts in livestock species preference or age profiles displayed in the bone assemblages.

While short-range transhumance (less than 10 miles) between summer and winter pastures is a well recognised feature of Scottish medieval livestock regimes, longer-range seasonal stock-movement – such as that from the Laich of Moray into the northern Cairngorm glens – is less well-known. Consequently, there is little understanding of how such long-distance movement was organised and regulated, and what archaeological imprint it might have left, e.g. in terms of overnight enclosures for animals, encampments or more permanent structures for the drovers, and other necessary infrastructure (dyked loans etc). A long-distance trade in cattle (and other livestock) is a well-known feature of the post-medieval agricultural economy in Scotland, but this droving trade can be shown from documentary records to have been functioning by the later 12^th century in some parts of the country and was an important dimension of royal estate-management in the 14^th and 15^th centuries, with the royal household being supplied with live cattle driven from pastures in Bute and fattened in the Torwood near Stirling. Monastic records in particular provide considerable insight into large-scale cattle management. Although Melrose Abbey, for example, has long been known as a wool-producer, its cattle operations have been less well recognised (Fawcett and Oram 2004). The abbey developed a complex of cattle farms, complete with byres capable of housing up to 120 animals, in the hills immediately north of Melrose, although, as yet no medieval structural remains have been identified in this area. The most important of its cattle-stations, however, was at Mauchline, where the abbey maintained an administrative and processing centre around which lay a series of satellite properties. From Mauchline, the herds were either disposed of through local burgh markets or were driven south down through Nithsdale and across the Solway fords for disposal at the Carlisle market. The abbey built up a chain of properties to act as overnight breaks on route or negotiated access to water or grazing at designated stopping-places. With the exception of the late medieval ‘Abbot’s Tower’ at Mauchline, nothing is known of the physical form of the main cattle station there or of any of the abbey’s overnight stopping-places (Fawcett and Oram 2004; Hall 2006). Valuable though better understanding of the form of such sites would be, perhaps of greater value would be the mapping through isotope analysis of the point of origin of the livestock being consumed.