‘Subsistence’ – what people ate in the past and how they acquired their food – has long been a primary concern of Mesolithic archaeology. This is sometimes mistakenly taken to imply that Mesolithic archaeologists have limited interest in issues about social organisation and ideology. That is not the case; subsistence behaviour is intimately connected to all aspects of lifestyle in past communities, especially those of hunter-gatherers. Its prioritisation is primarily one of methodological pragmatism rather than of theoretical persuasion. Indeed, it is important to recall that ‘subsistence’, ‘technology’, ‘social organisation’ and so forth constitute modern categories that are imposed onto past communities rather than these having any meaningful distinction in their own lives. A concern with what people ate simply provides a pathway into the holistic character of past behaviour and thought.
By definition, Mesolithic communities in Scotland relied on wild resources. The direct evidence for what specific plants were gathered and animals hunted is often sparse and in some areas non-existent. As a consequence, archaeologists frequently draw on comparative studies from contemporaneous cultures in regions with better preservation in Europe, such as Mesolithic southern Scandinavia, and with analogies from the ethnographic records of recently living and witnessed hunter-gatherers. In this work it is necessary to recognise that considerable diversity in diet is likely to have existed (for many reasons) across Europe in Mesolithic times just as in the present day.
It is easy to be sceptical about the value of ethnographic analogies because many of the recently documented hunter-gatherers were living in highly marginal areas, such as the San of the Kalahari (Lee and Devore 1976) and Central Desert Aborigines (Gould 1980), or were heavily influenced by contact with state societies and were using 'modern' technology, such as the Nunamiut (Binford 1978). Hence their relevance to Mesolithic communities in Scotland can appear at best tangential if not simply irrelevant. Nevertheless, ethnographic studies can provide an invaluable frame of reference for the study of prehistoric hunter-gatherers (Binford 2001). They must be used cautiously, but it is foolish to reject, a priori, any source of potential information and ideas when the challenge of reconstructing Mesolithic lifestyles is so demanding.
One of the main ethnographic contributions simply regards the scale of hunter-gatherer mobility. Most studies whether of San Bushmen in the Kalahari or the Inuit of the Arctic demonstrate one pervasive characteristic of hunter-gatherer lifestyles, that – with extremely few exceptions – they can cover vast distances in terms of their annual mobility patterns. This is neatly summarised by Lewis Binford’s statement that ‘archaeologists need to recalibrate their perspective of hunters and gatherers from the 5 foot square excavation unit at a single site to an area of more than 300,000 square kilometers’ (Binford 1983, 110). This was especially pertinent to Mesolithic Scotland where there had once been a focus on individual sites, for example the middens of Oronsay, as if these represented the entirety of the Mesolithic subsistence base. More recently, archaeologists studying the Mesolithic have attempted to undertake a regional approach, as in the Southern Hebrides Mesolithic Project, looking at the region of Islay, Jura, Oronsay, and Colonsay (Mithen 2000), and the Scotland's First Settlers Project looking at the Inner Sound of Skye (Hardy and Wickham-Jones 2009). While it may be that the spatial areas of even these projects only encompass a fraction of that covered by a single Mesolithic community -- Binford's suggested figure above would have covered the entire west coast of Scotland, from Arran in the south to the tip of Lewis in the north -- there are some caveats. It is necessary to remember that many of these studies encompass very different environments to those of post-glacial Scotland, and that other studies illustrate considerable variety and complexity in hunter-gather mobility. Perhaps the most important lesson is that mobility may just be a convenient modern catch-all for a complex pattern of movement in which different components varied in geographical scale, as they did in purpose, participants, and speed (see below).
Another invaluable insight from ethnographic studies concerns the patterns of mobility – or what is often termed a ‘settlement-subsistence’ system. In a seminal 1980 article, Binford differentiated between ‘logistically’ organised hunter-gatherers and what he termed ‘foragers’. The former organise themselves around residential base camps, normally occupied for a whole winter or summer season. From these, small groups travel considerable distances to ‘task-specific’ sites, locations which because of their situation were devoted to particular activities, such as watching for game, fishing, and collecting raw materials. Any resources acquired from these locations, sometime visited over a period of a few days, are returned to the residential base camp. Typically, different groups will be visiting different task-specific sites at the same time so that a range of resources are returned to the resident base for sharing. Binford contrasted this type of mobility pattern with what he termed foraging. This is where the whole group frequently moves their residential base and engages in searching the immediately surrounding landscape for resources without visiting any specific locations. This type of mobility pattern is suitable for relatively homogenous environments, whereas logistical mobility is appropriate for those which are more heterogeneous, in which different resources are found in quite different but predictable locations. These two types of hunter-gatherer mobility are the polar ends of a continuum. Any one community is likely to include aspects of both types of mobility pattern, perhaps switching between them in different seasons. As Binford explained, the nature of a hunter-gatherer archaeological record is strongly influenced by what type of mobility pattern is adopted.
An extreme form of logistical mobility is one in which the residential base becomes sufficiently permanent for the hunter-gatherers to be classified as sedentary. This has been used as a defining feature of so-called ‘complex hunter-gatherer’ groups (Price and Brown 1985). Without having to continually move their residential camp, hunter-gatherers are able to accumulate material items, maintain social hierarchies, build monumental structures and, perhaps, claim rights over land by acts such as the establishment of cemeteries. Whether any such complex hunter-gatherers existed in European prehistory is a moot point, but this notion has been used as an interpretation of the later Mesolithic archaeology -- the Ertebølle/ Ellerbek culture(Price 1985) -- from southern Scandinavia and Northern Germany. A prerequisite for hunter-gatherer sedentism is that there is a sufficient naturally replenishing source of wild animals and/or plants in the vicinity of the residential site. The most likely circumstance where this will occur is in coastal locations, which are supplied by a succession of migratory fish species, shellfish and crustaceans, birds, and maybe mammals, throughout large parts of the year, with the possibility of using stored food resources for the leaner seasons.
It is within this conceptual framework of potentially extensive mobility at varied scales, with the possibility of either logistical or foraging modes of organisation, or more likely combinations of both, and the possibility of ‘complex’ hunter-gatherer groups, that the Mesolithic archaeology of Scotland needs to be examined. This involves dealing, of course, with over 5000 years of human activity during which the climate and environment of Scotland went through significant change. Rather than imagining a single type of Mesolithic settlement-subsistence pattern, verying through the seasons, for this entire period, it should be seen as a continually changing evolution, as hunter-gatherers adapted to changing resource distributions and went through their own process of cultural evolution partly in response to independently conceived social pressures and partly due to environmental drivers. At present not enough is known about how people organised their time and how extensive their cyclic movements were. There are too many unknowns about the scales of mobility operating in Mesolithic Scotland to directly equate the evidence with ethnographic understandings of patterns of annual or even lifetime movements, let alone how these map on to theoretical models of subsistence-settlement patterns.
Estimating overall population size is based entirely on informed speculation based, again, on ethnographical input as well as statistical evaluation of the recovered evidence for Mesolithic activity in Scotland as a whole. From the figure of 62 people for the whole of Scotland in the Mesolithic envisaged fifty years ago (Atkinson 1962), perceptions of the period have moved to recent computer modelling of environmental productivity producing considerably larger numbers in the thousands (Tolan-Smith 2008). Determining the number, size, and composition of groups is, however, extremely difficult given the likely volatility of numbers over this very long period. Consequently it is to a different suite of issues around these topics that researchers begin to address with the archaeological evidence.
Understanding the human scale of life and how individuals articulated within wider social networks is also difficult to understand, especially given the paucity of actual human remains and the nature of the evidence. Consideration of the important issues of gender and childhood are equally difficult. At times, ethnographic evidence for gendered activity has been compared with Scottish archaeological material, for example the association of ‘Obanian’ middens with putatively female activities (Bonsall 1996). There are, however, alternatives to task differentiation models and traditional approaches to the sexual division of labour in relation to Mesolithic hunter-gatherers which acknowledge personhood and the fluidity of gendered identities across the life course (Finlay 2006). Consideration of age, individual personalities (e.g. Spikins 2008), and social skills helps to inform likely lifestyle narratives. Technological artefact analysis has also revealed different routines of production and subtle choices and distinctions in techniques which can be used to explore the social contexts of artefact creation and use (e.g. Finlay 2003; 2006; Warren 2006). These can be used to explore how the identities of both individuals and groups are experienced and expressed at different scales. The identification in some assemblages of novice stone workers, probably children, learning to knap (e.g. Coulerach, Islay; see Mithen and Finlay 2000; Finlay 2008) helps to restore humanity to lithic scatters and offers tangible pathways to addressing prehistoric knowledge acquisition and social values.
Mobility is a key element of our interpretation of the Mesolithic but it is hard to quantify archaeologically as this reconstruction set around the site at Kinloch on the island of Rum shows, © Alan Braby
The types of landscapes and resources that Mesolithic hunter-gatherers in Scotland were able to exploit were very diverse, including heavily indented coastlines with island archipelagos, coastal lowlands, rolling interior hills, large river valleys, and mountainous uplands. Aurochs, red deer, roe deer, wild boar, and otter were exploited, but less is known about the distribution of these animals in the landscape, especially regarding their presence on the smaller offshore islands. The Holocene woodlands are likely to have provided a diverse array of plant foods, while the coastal zone would have been especially productive with regard to sea mammals, fish, molluscs, crustaceans, seaweed, and birds.
It is indeed from coastal sites that the majority of direct information about Mesolithic subsistence derives, notably the midden sites on Oronsay, especially Cnoc Coig (Mellars 1978; 1987) and Sand, Applecross (Hardy and Wickham-Jones 2009) on the west coast and Morton (Coles 1971) on the east coast. Whether this is a true reflection of the significance of coastal resources within Mesolithic diet or simply a consequence of biased preservation and discovery remains unclear; evidence from elsewhere in NW Europe suggests that communities with predominantly inland or coastal territories may have co-existed (e.g. Schulting and Richards 2001). The middens of Mesolithic Scotland have often dominated archaeological approaches to this period, especially as a basis for the reconstruction of subsistence practice, and it is important to stress that there is still little understanding of the reasons for the construction of middens through repeated acts of deposition, nor of the cultural logics that made it appropriate for individuals to deposit certain kinds of materials in certain places (Warren 2007b). Archaeological orthodoxy often suggests that middens are a direct reflection of diet, whereas their relationship with subsistence strategy may be much more complex.
The Oronsay middens, dating to the early 4th millennium cal BC are regarded as later Mesolithic although they could possibly be early Neolithic, as farming economies are established elsewhere in Scotland at this time. Investigation of these sites has shown that limpets, periwinkles, and numerous other types of molluscs and crustaceans were exploited, while large quantities of saithe had been caught (see below for discussion of fishing technology). The mammal bones indicate the presence of otters, possibly hunted for their pelts as completely articulated skeletons were found, as well as grey seal, wild boar, and red deer. The grey seal bones were dominated by those of very young animals suggesting that hunting had occurred during, or shortly after the breeding season, which today is during September and October. The limited range of bone type of red deer and wild boar were primarily those used for making tools, rather than meat bones. It is highly unlikely that either wild boar or deer would have been present on the tiny island of Oronsay, or indeed, or its rather larger neighbour Colonsay, so either joints of meat were brought to Oronsay from further afield or maybe simply the bones themselves for tool-making. There were a large number of bird bones in the middens; although these have not been published, no less than 50 species were represented including shearwaters, lapwings, woodcocks, eider ducks, puffins, kittwakes, corncrakes, and swans – birds from a wide variety of habitats (Grigson, pers. comm.). Although no evidence has been reported, the eggs of these birds are also likely to have been an important food source.
Some of the most intriguing evidence from Oronsay comes from the otolith bones – the ‘ear bones’ of fish. The size of an otolith is a direct reflection of the size of the fish from which it derived. Because fish grow at a regular rate they can be used to determine the season at which the fish had been caught – assuming one knows the date of spawning. When Mellars and Wilkinson (1980) measured the otoliths from four different middens on Oronsay they found contrasting size distributions, which they interpreted as reflecting different seasons of fishing activity at each of the middens: Cnoc Sligeach in the mid-summer, Cnoc Coig in the autumn, Caistael nan Gillean II in early summer, and Priory Midden in the winter.
This evidence could be interpreted as Mesolithic hunter-gatherers inhabiting Oronsay all year round, moving from site to site with the seasons, presumably to avoid the worst of the prevailing winds and be close to the most productive resources (Mellars 1987). The isotopic analysis of human bone from the Oronsay middens (Richards and Mellars 1998) has also been used to infer settlement pattern. Samples from Cnoc Coig have indicated a diet with a very heavy reliance on protein from marine sources - which could be used to argue for a permanent presence on the island. But a sample from another midden, Caisteal nan Gillean II, has indicated a mixed diet of terrestrial and marine protein - which could be used to argue for seasonal movement between the coast and inland regions. Ultimately, the number of samples is too small to draw any firm conclusions. More likely is the alternative that the Mesolithic hunter-gatherers could have been intermittent visitors to the island. Oronsay being part of an extensive subsistence settlement system, this appears more likely in light of the tiny area and highly exposed nature of Oronsay. But until recently there was an absence of contemporary sites on the larger adjacent islands (Mithen 2000). Excavation of one contemporaneous shell midden site at Port Lobh on the west coast of Colonsay confirms that similar sites do exist and highlights that there is a need to be mindful of the biases created by the academic focus given to Oronsay (Finlay 2007a).
Combining the strands of seasonality and chronology, the sites in Oronsay have recently been interpreted as a response to economic stress in the later Mesolithic (Mellars 2004). This explanation is both plausible and convenient, while at the same time mirroring local historical information relating to the use of shellfish as a resource in times of famine (Wickham-Jones 2003). It is backed up by evidence that some earlier middens occur around what has come to be known as the 8.2k event (occurring c. 6200 cal BC), a brief but significant climatic downturn across NW Europe (Alley et al. 1997). However, while it is certainly possible that some middens do reflect the use of marine resources in times of hardship, this is unlikely to explain them all. And, if middens are a response to economic hardship, how prolonged were these episodes and how representative of Mesolithic Scotland as a whole?
What is unquestionable about the Oronsay evidence is the diverse range of coastal resources being exploited, or at least consumed. It was once thought that this type of coastal exploitation leading to the creation of middens was a feature of the final stages of the Mesolithic in Scotland, it being a key feature of the so-called 'Obanian', within which the midden from Risga was included. But the discovery of the midden at Sand (Hardy and Wickham-Jones 20092009), which dates to a relatively early stage of the Mesolithic, along with that of Ulva Cave (Russell et al. 1995) and An Corran on Skye (Saville et al. in press), suggests that coastal exploitation had been an aspect of Mesolithic subsistence throughout the period.
The coastal site of Fiskary Bay, Coll, stands in contrast with these midden sites, appearing to be focused on fishing alone (Mithen et al. 2007). Although a relatively small area has so far been excavated at this site, only fish bones have been recovered, along with wood charcoal, charred hazelnut shells, and chipped stone. The fish bones were only recovered by sieving excavated sediments through a fine mesh - they were otherwise not visible on the site - which is a, perhaps uncomfortable, reminder of the potential value of standardised recovery strategies as utilised elsewhere in Europe (e.g. Rensinck 2006). Although a similar range of coastal resources as represented in the Oronsay middens would have been available at Fiskary Bay, there was no trace of bones from sea mammals, land mammals birds, or molluscs. This may merely reflect the small area excavated as a diverse midden deposit might await discovery. Alternatively, Fiskary might be a specialised fishing camp in contrast to the generalised coastal foraging campsites represented by Cnoc Coig, Morton, and other midden sites. The bones come from a wide range of fish including wrasse, whiting, pollock, sea bass, and flat fish. Most of them appear to have been quite small, less than 30 cm, indicating a preference for inland waters where it seems most likely that they would have been caught by the use of fish traps. The location itself is ideally suited for this with a narrow inlet into the bay across which a wall can be easily built - a wall flooded at high tide but which would then trap the fish in the bay at low tide making them easy to collect in nets. The majority of fishing evidence from the Mesolithic of Scotland seems to indicate in-shore fishing, rather than deep sea fishing (Pickard and Bonsall 2004; but see Parks 2009). The value of the coastal zone in the Mesolihtic is also indicated by the recent excavation of fish weirs and fishing baskets in eastern Ireland (McQuade and O Donnell 2007; 2009; Mossop 2009), although none are known as yet from Scotland.
The Isle of Coll is unlikely to have had a population of red deer or wild boar during the Mesolithic and its woodlands are likely to have been relatively sparse (Wicks, forthcoming). Consequently, it seems likely that the island might have been visited only intermittently and for short periods, perhaps specifically for fishing at Fiskary Bay. The only plant foods evident from the excavation are hazelnuts, suggesting that at least some of the visits occurred during the autumn months.
While the quantity of charred hazelnut shell fragments at Fiskary Bay is limited, these have been found in vast numbers at the Mesolithic site of Staosnaig on the Isle of Colonsay (Mithen et al. 2001). This is also a coastal site, located in a sheltered bay on the east side of the island, but lacks any evidence for fishing, shellfish gathering, or exploitation of sea mammals. This may simply reflect the acidic nature of the soils at the site - the same type of sieving as used at Fiskary Bay had been employed but only tiny fragments survived within a unique micro-environment. A series of pits were found at Staosnaig surrounding a larger, shallow depression that contained a substantial quantity of charred hazelnut shell fragments - the remains of at least 40,000 nuts. This feature may have been the base of a hut which was subsequently re-used as a rubbish- or a cooking-pit.
Experimental roasting of nuts in pits similar to those found at Staosnaig has shown that if the shell becomes charred the kernel is inedible. However, it proved difficult to avoid about a fifth of nuts being charred during a roasting event (Score and Mithen 2000). So if those within the large depression were simply those discarded from the ovens represented by the surrounding pits, it implies that up to 200,000 nuts may have been roasted at Staosnaig. Quite why the remnants would be thrown into this depression - rather than just discarded on the ground or into the sea -is unclear; one possibility is that the nuts were being deliberately retained as fuel and burned to provide aromatic fumes for smoking fish or vegetables.
Resource gathering in Mesolithic Orkney, © Crown Copyright Historic Scotland reproduced courtesy of Historic Scotland. www.historicscotlandimages.gov.uk.
Whatever the precise number of nuts and the process by which they were placed into this depression, the evidence from Staosnaig suggests an intensive exploitation of the woodland on the island, especially because micromorphological analysis of the sediments indicates that the hazelnut shell rich deposit has accumulated very rapidly. Consequently, just as Mesolithic hunter-gatherers may have visited Coll specifically for fishing at Fiskary Bay, so they may also have gone to Colonsay specifically to harvest hazelnuts and, if so, that was not the only plant food being exploited on the island. Meticulous recovery and analysis has also identified the charred seeds and flesh from apples, the charred seeds of cleavers, charred seaweed, and charred tubers and 'bulbils' (small swellings at the intersection of the stem and leaves) of lesser celandine, a member of the buttercup family (Rannunculus ficaria). The frequency of the lesser celandine suggests that this plant had been deliberately gathered rather than entering the deposit accidentally. Ethnographic reports describe how it has been used as both a herbal flavouring for food and for medicinal purpose. The range of plant foods available wild in Britain, and the complexity of processing requirements for some of them, has been highlighted in the recent television series Wild Food (Mears and Hillman 2007). The development of methodologies for the identification of these resources, including, for example, starch and phytoliths, is a key priority for Mesolithic research and has profound implications for excavation and recovery strategies. Staosnaig and Fiskary Bay appear to be sites where subsistence activity is dominated by one particular activity -- harvesting hazelnuts and inshore fishing respectively -- in contrast to the much wider range of activities at the shell midden sites on Oronsay, and at Sand, Risga, and An Corran. More problematic to interpret with regard to subsistence are those Mesolithic sites which lack any faunal and floral material, although charcoal fragments and pieces of charred hazelnut shell are frequently present. Classic examples of these are those sites on Jura excavated by John Mercer which have high frequencies of microliths, such as Glenbatrick and Glengarrisdale (Mercer 1974; Mercer and Searight 1986), Starr and Smittons in the south west (Edwards 1996) and those on Islay such as Gleann Mor (Mithen and Finlayson 2000) and Bolsay located on the Rinns peninsula (Mithen et al. 2000).
For sites with limited direct evidence for subsistence, indirect evidence such as their location in the landscape, and the structure of their tool assemblages is used to construct hypotheses about their place in wider systems. Sites do vary in the size and character of the assemblage from small apparently discrete, sites like Fife Ness, near Crail (Wickham-Jones and Dalland 1998), which may represent a short temporal event, to larger palimpsests which probably highlight the use of recurrently favoured locations in the landscape, e.g. at Kinloch, Rum (Wickham-Jones 1986). In this respect some of the Islay sites are of particular interest because they are inland - although only a few kilometres from the coast. The excavation at Bolsay produced more than 5000 microliths from an area estimated to be less than a fifth of the spatial extent of the artefact distribution. The traditional view is that such artefacts were being manufactured as the points and barbs for arrows with which game, notably deer, were hunted. Although microwear analysis has shown that some microliths were used as drills and knife blades, others have impact fractures suggesting that there were indeed used as hunting weapons (Finlayson and Mithen 2000). As well as being present in particularly large numbers, the microliths at Bolsay show far less diversity in form than at many others sites, suggesting specialised activities. Moreover, the rarity of features and lack of spatial patterning in the artefact distributions suggest that the large numbers of artefacts had accumulated through many short-term visits to the site. Taking all of this into account, along with the fact that the Rinns of Islay remains today as a favoured location for hunting deer, it seems likely that Bolsay may have been another task-specific site - one for hunting red deer. Similar interpretations are likely to apply to the microlith-rich sites on Jura and elsewhere in Scotland, although some of these, such as Kinloch on Rum, have more diverse artefact assemblages suggesting a wider range of activities (Wickham-Jones 1990). At the Sands of Forvie, Aberdeenshire, large lithic assemblages suggest an emphasis on the manufacture of microlithic components and, possibly, the extensive exploitation of starchy plants; one interpretation is that this site was utilised for preparing tools and cord/string, for use elsewhere.
Some of these microlith-rich sites are located adjacent to inland rivers, suggesting that fishing for salmon may have been the primary subsistence activity. Most notable are sites such as Rink Farm located at the juncture of the Rivers Tweed and Ettrick Water. Sites like Dryburgh Mains and Kalemouth (Callander 1927; Mason 1931; see also Mulholland 1970; Warren 2005), are all located at key locations today for salmon fishing, and Nether Mills is adjacent to the River Dee (Kenworthy 1981). Problems exist in such interpretations; little is known about salmon behaviour in the early Holocene, especially the predictability of runs in the medium term, and uniformitarian analogies with modern salmon-run timing and abundance are not appropriate (Warren 2005, 57-58). This highlights the over-arching problem of reconstructing past environments for which modern analogies do not exist (Spikins 1999).
Many Mesolithic sites in Scotland lie on or near the coast but it is pertinent to ask whether this is evidence of a true preference for coastal living, or simply an archaeological bias. The latter is most likely, for while Mesolithic sites in the interior are few there has not been the same tradition of research activity (Finlay et al. 2002; Finlay in press). Sites such as Chest of Dee, Aberdeenshire (Murray et al.2009) and Ben Lawers, Perthshire (Lelong 2003) indicate that Mesolithic activity penetrated the Highland Zone, and Daer Reservoir, South Lanarkshire (Ward 1995; 1997; 2000a; 2010) and Starr on Loch Doon, Dumfries and Galloway (Affleck 1986) provide evidence in other eco-zones away from the sea. A number of factors that do not operate in the interior have facilitated the discovery of sites in the coastal zone (Wickham-Jones 2004b): active and large scale erosion has worked to reveal sites; the working of agricultural land has resulted in the recognition of lithic scatter sites (especially in the decades preceding the mechanisation of farm machinery); developments such as road workings and building projects have uncovered material; and the emphasis of population today on these areas as opposed to the higher interior land has provided both a community with some interest in archaeology and a focus for development such as that mentioned above. In addition, groundcover in the interior has often militated against the recovery of sites, especially blanket forestry or peat; while the sheer inaccessibility of much of the zone for the population of today has reduced present activity and thus recognition of archaeological remains; this latter point has been reinforced by the systems of land holdings and use over much of the interior. Mesolithic sites in the Highland zone are more or less confined to land belonging to a single landowner with an archaeological interest. Finally, it should be noted that archaeology, as a profession, has contributed to the coastal bias in that the siting of a few large-scale research projects especially on the west coast has both raised awareness among local people to the extent that new sites are recognised, and stimulated further research questions among archaeologists who have then moved into the area with a virtuous circle of yet more projects.
The lack of detailed work in the interior makes it hard to quantify the relative importance of the coast to the inhabitants of Mesolithic Scotland but this is a pattern that has begun to change especially within the last decade or so. The concentration of sites does suggest that the coastal zone may well have formed a preferred location for many activities, at least for many groups. Nevertheless, it is also relevant to query whether it was the coast that was of value or the wider lowlands, many of which lie within easy reach of the coast. The concentration of sites along rivers such as the Dee or the Tweed should be examined in more detail and put into a wider perspective. While the coast offered many advantages: a wide, year round, resource base for foodstuffs; abundant fresh water, easier mobility; lithic materials; shelter; and a lack of insects - many of these features were also to be found along the river valleys.
The relationship between Mesolithic communities and the resources they exploited has long been debated. The deliberate introduction of ungulates to islands such as Shetland is claimed and may indicate a more complex relationship between humans and large mammals than is implied by the idea of 'wild' game and a long interpretative tradition suggests that Mesolithic communities managed their landscapes by deliberate firing (Mellars 1976; Simmons 1996), in order to promote browse. This is not as well demonstrated in Scotland, and the best evidence here comes from the south. It may be that the very different environments of northern Scotland required different strategies. Although no actual remains have been found, it can be assumed that the Mesolithic hunter-gatherers in Scotland had domesticated, or semi-domesticated, camp dogs which provided assistance in hunting, protection, and companionship (Munt and Meiklejohn 2007).
Mesolithic hunter-gatherers needed to acquire more than just food from their landscapes. Firewood and material for making stone artefacts would have also been essential and their collection is likely to have been embedded into other routines of hunting and plant gathering, or vice versa. The site of Coulererach, also in the Rinns of Islay, is adjacent to what is today one of the most flint-rich beaches in the whole of western Scotland being closest to the Ulster source of this material in the glens of Antrim (Mithen and Finlay 2000). Not surprisingly, therefore, chipped stone assemblages are dominated by the very initial stages of pebble reduction: many tested and then discarded pebbles, large numbers of primary flakes, and numerous unworked pebbles. Whether Coulererach was another task-specific site – raw material acquisition and preparation of cores for transport – or had also been used for generalised coastal foraging activities remains unclear because the peaty acidic soil has destroyed any bone or shell material that may have been present.
Despite the long period covered by the Mesolithic in Scotland, and the fragmentary nature of the evidence, and hence, interpretation, the picture that emerges is generally one of a settlement system that tends towards the logistical mode of organisation in Binford’s terms. This is not surprising in light of the environmental diversity of Holocene Scotland and hence the considerable seasonal and spatial variation in the distribution of resources, even before consideration of the extensive change over time that must have characterised these distributions given the dynamic nature of climate change throughout the period. But this does not imply any fixed, rigidly cyclic nature for Mesolithic mobility. All hunter-gatherers have to be flexible to respond to unexpected events such as the beaching of whales, providing a sudden glut of food and raw materials, poor summers for ripening fruit, a failure of the seabirds, and, of course, environmental disasters like the Storegga Slide tsunami, would all have had an impact on the annual routine. Several sites in the Firth of Forth include the bones of whales associated with Mesolithic artefacts – at Meiklewood an antler mattock was possibly left propped against a rorqual whale skull (Warren 2005, 120). Other sites, representing hardship, may be more difficult to identify, but, as indicated above, some studies have examined the possibility that some midden sites may indicate a response to famine conditions (Woodman 2001).
Indeed, one of the most important preoccupations for any hunter-gatherer is simply to observe and know the landscape: the sky, sea, flight of birds, and so forth, from which information can be extracted about where resources might be located. One further site on Islay, Aoradh, has been interpreted as used for precisely this reason - an observation camp - it being located adjacent to the location of a RSPB hide used by modern-day ornithologists (Mithen et al. 2000), while GIS analysis has shown that Mesolithic sites on Islay are located with more extensive viewsheds than one would expect by chance alone (Lake et al. 1998).
While it is tempting to infer a logistical pattern of mobility for Mesolithic hunter-gatherers, actually documenting the scale of their mobility is extremely difficult. It is also important to remember that mobility does not necessarily mean that the whole community is mobile, or that mobility takes place year round. Even if there were sites with identical radiocarbon dates from, say, Skye and Arran, it could not necessarily be concluded that these were created by the same community of highly mobile hunter-gatherers, rather than two distinct communities. One possible means of directly tracing mobility is by considering the distribution of raw materials from known sources (see section 5.5). Pitchstone from Arran, for instance, is found at Bolsay on Islay and at Staosnaig on Colonsay, most likely having been directly carried there rather than as a product of down-the-line exchange (Mithen 2000). Similarly, bloodstone from Rum has been found at Mesolithic sites on Ardnumurchan and Morven at distances of c. 50Km from source (Clarke and Griffiths 1990). But other sites at an equivalent, or indeed shorter, distance, such as Fiskary Bay on Coll have no traces of bloodstone within its raw materials, which in this case are entirely of flint. In general, it appears that Mesolithic hunter-gatherers probably made use of whatever raw materials were locally available and consequently, using the distribution of any particular raw material type to document patterns of movement or to support the interpretation of extensive exchange networks is open to alternative explanation. Work elsewhere in Europe suggests that geographical distributions of particular materials and particular object-types exist, but that the extent of these overlapping distributions does not always match, implying that caution is required in using any single type of evidence to discuss the scale or extent of any proposed movement (see Bergsvik 2003).
Whatever the extent of Mesolithic mobility, it is unquestionable that activity occurred on islands that required the crossing of at least 20 Km of open and often turbulent water. Unfortunately there is currently no trace of the boats used. Those from Mesolithic Denmark tend to be made from hollowed-out large tree trunks, especially of lime, such as from Tybrind Vig (Andersen 1985) and similar vessels may have been used in some parts of Scotland. Whether trees of equivalent size and quality were available is a moot point and, given the availability of experimental evidence for open sea crossings in hide - covered wooden framed boats (Severin 1978), the use of coracles and larger, easily beached and manoeuvred hide-covered boats is likely. Many of the skills and techniques used to construct these craft would also likely have been used to construct shelters and other structures. The evidence for ground-fast structures (Wickham-Jones 2004; and section 4.1.2 ) reveals a diverse array of pits, small , putatively roofed enclosures, and stake and post-hole alignments suggestive of structures and other installations (like drying racks). Understanding how these features were built, what they actually looked like, and even what materials were used for their construction, is, however, more problematic. It is here that more experimental work based on archaeological results would help to give substance to the appearance of sites and would allow reconsideration of the longevity and resilience of these seemingly ephemeral structures.
Used lump of ochre from the excavations at Sand © Scotland's First Settlers
Questions can also be asked about the costume and personal equipment and adornment of these Mesolithic people. Direct evidence for clothing is lacking, and although this is often inferred from the presence of scrapers and microscopic traces thereon of hide-working it cannot be assumed to have been present. Some evidence demonstrates the use of pigments; ochre 'pencils' for example, were present at Morton (Coles 1971). Traces of distinctive minerals in pit fills at Warren Field, Crathes, imply the exploitation of spectacularly coloured purple and green rock outcrops at the Pass of Ballater some 40km from the site (Murray et al.2009) and lumps of similar colourful minerals have been found at other sites (e.g. Mithen and Finlay 2000). Perforated cowrie shells have been recovered from Carding Mill Bay, Sand, Ulva, and several of the Oronsay sites (see Saville 2004c,200-202 These finds hint at personal decoration and adornment such as the use of bead decoration (Simpson 1996), but the finds are too fragmentary to allow for any meaningful interpretation. There is an absence of the shale and stone beads and tooth and amber pendants found in Mesolithic contexts elsewhere in NW Europe.
Evidence to inform any understanding of the ritual practices and religion of Mesolithic communities in Scotland is extremely limited. Elsewhere in northern Europe, naturalistic and shamanic beliefs are seen as dominant motifs of Mesolithic lifeways (e.g. Schmidt 2000; Zvelebil 2003a). Among modern groups these beliefs result in particular practices and are expressed in material culture (Jordan 2001). Global perspectives on recent hunter-gatherer ritual and religion reveal considerable diversity and complexity in ritual practices (see papers in Lee and Daly 2000). These highlight the importance of distinctive landscape features, seasonal events, attitudes towards discard and depositional practices, and the treatment of the dead. With these themes, the meanings of particular places, cultural attitudes towards particular resources, discard, and the dead in these communities in Scotland can be tentatively explored. Regional distinctions and the long duration of the period need to be acknowledged, however, and how understandings of land and seascapes as well as beliefs and practices would change.
No burials are known from this period in Scotland, although this may be a reflection of excavation bias such as the paucity of cave excavations under modern conditions. To date discussion of burial is restricted to the bones found within the Oronsay middens, mainly Cnoc Coig. These are predominately hand and foot bones. One suggestion has been that these are the remnants of bodies that had been laid across the midden to decay as part of a funerary process and once de-fleshed taken elsewhere for burial, leaving some of the smaller bones behind (Pollard 1996). Some of the bones, however, appear to form distinct clusters and because these are composed by those from more than one individual, seem to have been deliberately positioned in that manner (Meiklejohn et al. 2005). Moreover, one of these clusters contained the flipper bones from a seal. It is, of course, tempting to read significance into that in light of the frequent role that seals play in Scottish oral traditions in which they are often thought to represent either the souls of the damned, the bewitched, or the reincarnation of those lost at sea. Alternatively, one might suppose that this deposit indicates some kind of link being made between the human and the animal, and this reflects a widespread characteristic of hunter-gatherer cosmology - that animal and human are not seen as sharply differentiated (Conneller 2009).
It would be unsurprising if the Mesolithic hunter-gatherers had not imbued the striking topographic features and dynamic geomorphology of the Scottish landscapes with mythological significance, perhaps creation myths as argued for elsewhere (e.g. Zvelebil 2003b). With regard to the west of Scotland one need only think of the impact that, say, Fingal's Cave of Staffa, the Paps of Jura, or the Sgurr of Eigg makes on people today, and the wealth of stories associated with these landmarks, to easily imagine that they would have played a role in the Mesolithic cultural interpretation of their landscapes. Evidence from Warren Field, Crathes, Aberdeenshire, implies the exploitation of a strikingly coloured rock outcrop (Murray et al.2009). Manor Bridge and the Dookits, both on the Upper Tweed Valley, are located near rocky outcrops immediately adjacent to the River Tweed (Warren 2005). These are far from dominant features, but are locally important and likely indicators of the possible integration of routines of movement and natural features of the landscape.
Art appears to be absent; no figurative art is present on the Scottish bone and antler implements in contrast with other areas of Europe and no incised or decorated pebbles such as those recovered at Rhuddlan, Wales (David and Walker 2004), are yet known from Scotland. During this period symbolic expression may have taken different material forms. A Mesolithic pit alignment at Warren Field, Crathes, is argued to have had a symbolic role, with the deliberate deposition of materials, including food stuffs (Murray et al. 2009). The potential role of middens as Mesolithic monuments has seen a great deal of recent attention and has been strongly critiqued (Warren 2007). Rather than assuming that they are a 'monument' (especially as most middens are not upstanding) a more detailed consideration of the architecture of middens, including the other structures they include, would be a very important contribution to this debate. The movement and redeposition of midden material, and possible associations with burning does appear to be significant more generally in the Mesolithic (e.g. Newton, Islay; McCullagh 1989). Therefore more attention needs to be given to the nature of discard practices and other strategies and exploring how cosmologies and beliefs may be expressed in daily routines.
Of course many significant events in modern hunter-gatherer life such as those marking rites of passage, acts of initiation and healing trances will leave few archaeological signatures but this should not mean that they can be ignored and there are some ways of indirectly approaching these subjects such as looking for periods of ritual exclusion expressed as differential growth markers in teeth. Equally, a more creative and imaginative approach, exemplified recently in The Gathering Night, a fictional work by Margaret Elphinstone (2009), presents arhaeologists with the challenge to seek answers to other, often difficult, questions about belief and ritual practice. Fundamentally, however, knowledge about the lifestyles of Mesolithic peoples (and more so for earlier Palaeolithic groups) is highly constrained by the evidence. It is therefore the smallest of traces and such momentary encounters that must be employed to enrich the narratives of their lives in the land and islands now known as Scotland.