In This Section:

4.1 Introduction

  1. The Upper Palaeolithic and Mesolithic periods cover a huge span of time, which extends from the arrival of the first humans in our region from roughly 12,500 BC to the widespread adoption of agriculture around 4000 BC. South West Scotland has similar patterns of occupation and activity to those evident throughout northern Britain. 
  1. The evidence suggests small groups, probably familial in structure, living as hunter-gatherers, having seasonal ranges they moved around with very limited durable settlement. There is evidence for only a transitory human presence in the Upper Palaeolithic (Old Stone Age) during the Late Glacial period. More substantial evidence for occupation of our region comes from after the last Ice Age, in the Holocene, during the Mesolithic (Middle Stone Age). Throughout this later period, mobile groups survived through gathering, hunting and fishing within a landscape that was being transformed by environmental change (see Chapter 3). The predominant evidence used to locate their sites is the debris from manufacturing flaked stone tools – known as lithic scatters. 
Landscape plan of South West Scotland showing topography, rivers and coastline. The plan is overmarked with the location of Key Knowledge Sites.
Distribution map of Key Knowledge Sites in the SWSARF region © Rathmell Archaeology Ltd
andscape plan of South West Scotland showing topography, rivers and coastline. The plan is overmarked with the location of Other Named Sites.
Distribution map of Other Named Sites in the SWSARF region © Rathmell Archaeology Ltd
  1. South West Scotland has played a crucial role in developing our understanding of Mesolithic Scotland. Initial interest was sparked by local antiquarians like John Smith of Dalry and William Downs of Irvine. Their explorations recovered lithic assemblages that would later be termed ‘pygmy flints’ at Ardeer, Shewalton and elsewhere in the region (Smith 1882; Smith 1893; and Smith 1895). Similar diminutive forms of lithic implement (later termed ‘microliths’) were also identified at Glenluce Sands by Ludovic Mann and John Callander (Mann 1911, Callander 1911), and by John MacArthur at Black Cave, Arran (MacArthur 1861). These microlithics were recognised as some of the earliest tool forms being elements for composite tools. 
  1. Not all of the early discoveries in our region were of stone tools – two red deer antler harpoons were recovered from rivers. In 1895 at Cumstoun a harpoon was found on the bed of the River Dee (Munro 1908), then in 1938 another was discovered on the bed of the River Irvine at Shewalton (Lacaille 1939).  
  1. These concentrations of lithics, including microliths, and limited antler tools, attracted the attention of Arnaud Lacaille. He initially studied the material from dune systems and raised beaches at Stevenston Sands, Shewalton, Glenluce Sands and Ballantrae. At Ballantrae during the 1920s-1930s some 3,000 lithic pieces, including microliths, had been collected from ploughed fields by William Edgar (Edgar 1939, Lacaille 1945). Lacaille’s initial work, which included the first use of the term ‘Mesolithic’ in Scotland (Lacaille 1930), was followed by a broader analysis of Stone Age artefacts throughout Scotland (Lacaille 1954). Lacaille’s research created the benchmark which defined the study of Mesolithic Scotland for at least four decades. His division, into an earlier ‘Larnian’ phase (with its implied affiliation with the north of Ireland) and a later ‘Obanian’ phase, would, however, be challenged. 
  1. It may have been Lacaille’s focus on South West Scotland which, over the following decades, encouraged others to continue the search for evidence of Mesolithic activity in the area. Certainly, the work of Lacaille promptly informed the recognition of a Mesolithic lithic assemblage that was recovered at Terally during the excavation of a series of long cists (Livens 1956) and the recovery of more material from Glenluce Sands (Davidson 1952, McInnes 1964) on the raised beaches. 
  1. The 1960s and 1970s saw fieldwork by local archaeologists Michael Ansell and William Fleming Cormack. Between 1962-4, Cormack discovered 17 new sites mainly around Luce Bay, but also around Burrowhead and Loch Ryan, and further east towards Kirkcudbright. Surface collections by Malcolm MacNeill between 1965 and 1976 also added numerous coastal site between Ballantrae and West Kilbride on the Ayrshire coast, with large collections from Girvan (Morrison 1981). This reinforced the coastal distribution of known sites, associated with the raised shorelines which mirrored the availability of free flint on the modern foreshore and in raised beach deposits (Gemmell and Kesel 1977). 
  1. These sites were identified as lithic scatters – dense concentrations of worked stone objects and waste material from their manufacture revealed by modern disturbance. Sometimes this material was revealed due to ploughing arable ground, or forestry operations, or eroding stream channels, reservoir margins and peat hags. Often, these sites had been heavily disturbed by later activities, while the scatters also included artefacts from later periods such as the Neolithic and Bronze Age. But in each case, it required researchers to be actively surveying the landscape at the time when lithic material was exposed (Barrowman 2000 explores research of lithic scatters in depth). 
  1. Work continued on the analysis of lithics, and this ongoing study did much to challenge Lacaille’s original assumptions. Coles argued instead for a South West Coastal Mesolithic, where the local industries were influenced in broad terms by northeast Ireland but also by other areas, in particular northeast England (Coles 1964). The subdivision between ‘Larnian’ and ‘Obanian’ cultures was replaced by the recognition of an earlier broad-bladed microlithic industry and a later narrow-bladed industry. The broad blade assemblage included microliths that were isosceles triangles and obliquely blunted points. In contrast, the narrow blade assemblages – which were more common – included microliths that were scalene triangles and crescents as well as edge-blunted forms. 
  1. Cormack went on to excavate two coastal lithic scatters on Luce Bay – Barsalloch and Low Clone – revealing stratified deposits related to contemporary Mesolithic occupation. These sites were excavated from 1965-7 by Cormack and enabled the first radiocarbon dates with which to anchor typologies of lithic tool forms. Both sites were interpreted as settlement sites, although the scale of features present varied. Low Clone had a large flat-bottomed hollow associated with hearths, stakeholes and stone settings, while Barsalloch had only evidence of hearths and stone settings.  
  1. At around the same time, new sites were also being found further inland, in upland locations around the Rivers Ken and Dee. This work may have been informed by Truckell’s advocacy in a review of the Mesolithic of the region for survey at the top of river terraces of south-flowing rivers to identify occupation of inland areas (Truckell 1963). 
  1. The region’s first inland site was identified in 1967 by Ansell at Kendoon, in a riverine location at a height of around 100m OD, closely followed by a series of sites around Loch Doon (Ansell 1968, 1969), including Starr and Smittons. Within a few decades, the number of recorded sites had risen to 50, all located within the inland catchments of the River Doon, Black Water of Dee and Water of Ken (Edwards et al 1983). It was soon argued that their distribution potentially illustrated the use of river valleys through the uplands linking the Ayrshire Coast and the Solway in addition to exploiting upland resources, including chert deposits for tool manufacture. 
  1. These upland sites would not start to be investigated until the 1980s, when Starr and Smittons were excavated by Tom Affleck in the period 1983-6 along with Auchareoch on Arran. These excavations were in association with Kevin Edwards and Ann Clark, continuing a focus on these upland sites started by Ansell and with Allen on Arran (Allen and Edwards 1987). All three sites showed evidence of hearths, stakeholes and occasional pits co-located with the lithic scatter that led to the identification of each site. The ephemeral nature of the structures and activity areas reinforced interpretations of temporary, transient occupation of the landscape. Neither Starr or Smittons are fully published, though both have considerable bodies of published material linked to them. 
  1. Coastal work continued, with lithic collections from Auchareoch on Arran (Fairhurst 1981), more lithic recovery from Glenluce Sands (Idle and Martin 1977), while Lionel Masters in 1977 excavated a hearth within an oval pit at Redkirk Point, near Gretna. This hearth site had no artefacts, but a wealth of environmental evidence.  
  1. During the 1970s, Paul Mellars used statistical analysis to create classification schemes of Mesolithic settlement types and worked stone assemblages (Mellars 1976). One of the Luce Bay sites – Low Clone – was classed by Mellars as a Type 2 settlement on account of its large size (measuring between 44 and 210m2 in extent). Its Mesolithic lithic assemblage was also interpreted by the same author as a Type B ‘balanced flint assemblage’ on account of its character and composition, which consisted of 30-60% of microliths and 25-50% scrapers.  
  1. Alex Morrison provides a series of synthesis papers covering the Mesolithic of South West Scotland in the aftermath of these discoveries (Morrison 1981, Morrison 1982). Within this, he advocates that undated human burials from Ballantrae (Moore and Smith 1885) may be Mesolithic, drawing links to the Danish Mesolithic in the burial rite and the Ballantrae environment. In seeking to establish an overarching chronology for Mesolithic occupation, Morrison recognises that a framework was available from sea-level change and coastal development. These synthesis papers rely on this chronology given the scarcity of absolute dates from these sites. 
  1. All of the excavations yielded radiocarbon dates, which at last provided a means of locking South West Scotland’s Mesolithic material into an absolute dating framework. Nuances in the data were already becoming apparent: the hearth at Redkirk Point (7100 to 6500 BC), for example, yielded Mesolithic dates earlier than those recovered from the coastal site of Barsalloch (5200 to 4600 BC). Auchareoch (7300 to 6000 BC) proved to hold the earliest evidence comparable to Redkirk Point, while inland Starr (5500 to 5000 BC) and Smittons (5500 to 4100 BC) were closer in date to the Luce Bay coastal sites. Evidence of the subsistence basis of the communities was also emerging with marine shell recovered from Barsalloch, red deer from Low Clone and hazelnut shells from Smittons and Auchareoch. 
  1. The importance of the Mesolithic sites at Auchareoch, Barsalloch and Smittons was highlighted by their inclusion in a review of the early Flandrian settlement in Scotland, presented at the 1985 international conference on The Mesolithic in Europe (Morrison and Bonsall 1989). At this point the region provided a quarter of the absolute dated late Mesolithic sites in Scotland. Another important aspect of the paper was an acknowledgement that residual Upper Palaeolithic material may be present within larger assemblages broadly characterised as Mesolithic. They pointed out that there was no reason why human activity could not have commenced in the region early in the Late Glacial period, or even slightly earlier. 
  1. Research by Isobel Hughes into the origin of Neolithic communities in the region (Chapter 5) led to an appraisal of the Mesolithic in the region (Hughes 1988). This appraisal relied heavily on Morrison’s synthesis papers to argue for ‘complex’ coastal hunter-gatherer societies that gradually transitioned to settled agriculture, expanding into the upland areas, marking this through chambered cairns. Robert Chapman challenged some elements of the appraisal on the grounds of a lack of evidence for sedentarism and the danger of simplifying the cultural variability in Mesolithic communities (Chapman 1988). This debate was later revisited by Jane Murray (Murray 1991) who explored the co-location of chambered cairns with earlier lithic scatters on the Solway coast; though Hughes’ response is principally focused on Neolithic issues (Hughes 1991), there is a strong rebuttal of the proposal that navigation across and between the islands of the Clyde would pose a conceptual barrier. 
  1. During the second half of the 20th century, a wealth of environmental data was also accumulated, sometimes as an adjunct to archaeological works but often as a focus for study in its own right (Morrison 2010). This research started building a palaeoenvironmental history for the region that included the Mesolithic (Robinson 1981, Newell 1990). Work on the raised beach deposits along the Clyde and Solway coasts (Boyd 1982) also developed a richer understanding of the changing seascape (Chapter 3). This research was often interwoven with consideration of how early human occupation could utilise and live within the landscape (Edwards 1989).  
  1. From the late 1980s onwards, the number of local and academic researchers investigating the Mesolithic in South West Scotland, as opposed to using data in synthesis and academic debate, dwindled. This was in part due to the premature death of Tom Affleck (Edwards 1996). Regardless, the focus in Scotland shifted towards the Hebrides, which dominated the focus of academic researchers to the end of the century and beyond – for instance, Paul Mellars (Oronsay), Clive Bonsall (Oban and Ulva), Steven Mithen (Islay) and Caroline Wickham-Jones (Rhum). 
  1. This shift impacted doctoral work with, for instance, research on lithic raw materials and settlement structure (Marshall 1997) examined an area of western Scotland that only drew Arran into its remit, which was otherwise looking at the Inner Hebrides. An exception to this trend was the doctoral research by Wright (Wright 2012), whose focus on west central Scotland encompassed Ayrshire with particular attention to Ballantrae and Girvan. The drive to date museum holdings of early evidence for human occupation in Scotland benefited the Cumstoun (5700 to 5500 BC) and Shewalton (4900 to 4500 BC) harpoons (Bonsall and Smith 1989, Bonsall and Smith 1990). 
  1. While residual Mesolithic material would be revealed in academic research focused on other periods like Kilpatrick (Barber 1997), Machrie Moore (Haggarty 1991) and Holywood North (Thomas 2007b), from this point forward, purposeful research would be driven by developer-led work. Researchers in commercial practice are constrained to mitigating the development impacts their clients generate – they do not directly choose where in the landscape they work. However, commercial practice assumes that the absence of archaeological sites at a location must be tested, not accepted. This was a fundamental break with earlier researchers (Chapter 2), which has led to an increase of Mesolithic sites investigated with different associated features being revealed. This research approach is partial, just as earlier local and academic research was; the research is focused on where development is commercially viable and consented to. Further, the successful completion of the research can be conditional on the management of the larger development project. 
  1. Some of the more significant discoveries have been revealed by linear projects that intersect with numerous sites, both known and previously unknown. In particular, the progressive upgrading and development along the A74/M74 corridor running up Annandale, the western end of the A75 corridor along the Solway coast, and the Arran Ring Main have added to our known sites during the 1990s and 2000s. 
  1. The construction of the Arran Ring Main in 1999-2000 (Donnelly et al 2000, MacGregor 2000) was a rich source of information on the Mesolithic occupation of the island, though this work has not yet reached publication. At Monamore Bridge, a series of buried occupation horizons enclosed within a dune bunker from the late Mesolithic to the late Neolithic was revealed with associated pits, isolated postholes and a possible occupation scoop which may represent early Mesolithic occupation. A mixed date lithic scatter was also excavated at Bridge Farm (Baker 2000), which has also not reached publication.  
  1. In other parts of the region, lithic scatters without notable accompanying features have been excavated at Tarf Water and Kirkhill Farm (Pollard 1993). The latter was on what had once been the bank of an old river channel, marking a former course of the River Annan.  The former, excavated in 1995, though substantial in size, has not reached publication. 
  1. In some instances, a Mesolithic presence has been attested by the recovery of occasional residual artefacts (usually microliths) such as at Beckton Farm (Pollard 1997). The interplay between Mesolithic material and the later site can spark a review of whether the material is residual. An unusual lithic object with microlithic tendencies at Beckton Farm led the excavator to argue that it may show the retention of Mesolithic stone-working techniques within a later, Neolithic assemblage (the focus of that site). 
  1. Elsewhere, attribution of an artefact-free feature to the Mesolithic has been achieved through radiocarbon dating at Evan Road for a hearth (6000 to 5800 BC). Residual burnt hazelnut shell fragments have also provided the earliest dating evidence for activity from within a scoop (5000 to 4800 BC) at Hunterston (Gordon and Turner, forthcoming) and under burnt mounds at (4500 to 4300 BC) at Hallmeadow (Green 2022). The base of an urban excavation in Dumfries at Irish Street revealed a very limited lithic assemblage with a wind break or drying rack, while Mesolithic burnt material was recovered from a gully at Tralong Wind Farm (Gray 2021). These discoveries highlight that not all Mesolithic activity is co-located with diagnostic lithic material.  
  1. Mitigation of multi-period sites has increasingly encountered fragmentary or residual evidence for Mesolithic activity. Excavation in 2011 of a multi-period site in Monkton (Rennie 2015) recovered a small lithic assemblage associated with a pit cluster; no dating evidence was recovered. Earlier evaluation of a nearby development site between 1997 and 2000 at Monkton also examined a Mesolithic lithic scatter (Cameron 2000). A small pit cluster and very limited lithic scatter was excavated at Ayr Cemetery (Paton 2018) associated with hazelnut shell fragments (5900 to 5750 BC and 4490 to 4360 BC). Excavation of the Bronze Age roundhouse on the raised beach at Ross Bay (Ronan and Higgins 2005) recovered a small lithic assemblage. Investigation of multiple sites at Laigh Newton (Toolis 2011) recovered a residual lithic assemblage that may be either late Mesolithic or early Neolithic, residual charcoal (6400 to 6200 BC), as well as hazelnut shell-rich pits (3650 to 3500 BC). Evaluation work at South Boreland revealed numerous prehistoric features, including a lithic-rich shallow deposit that appears to be late Mesolithic (Engl and Wilson 2015).
  1. Girvan has also proved to be a hotspot for discoveries. Mitigation of sewer construction in 1996-8 at Gallow Hill revealed several pits (4800 to 4500 BC) associated with one of two Mesolithic lithic scatters of around 1400 pieces. The analysis of this multi-phase assemblage provides insights into the Tarf Water assemblage, amongst others from across the region. To the immediate northwest was Littlehill Bridge, where construction in 1994 revealed a scooped occupation hollow (6400 to 6000 BC) with a lithic assemblage.  
  1. Residential development in Barassie from 2002-12 revealed pits, hearths and large volumes of hazelnut shell, suggesting a dedicated food processing site (7300 to 7000BC). Excavations for the new Ayr Academy (Arabaolaza 2019) revealed a posthole and a pit cluster (5200 to 5000 BC) as well as a varied lithic assemblage that included both late Mesolithic narrow blade and early Mesolithic broad blade material. 
  1. Developer-led work to the south of Ayr for residential development also revealed several previously unknown settlement sites. Fieldwalking, test pitting, and excavation at Ailsa View recovered a substantial lithic assemblage of over 3000 pieces from a ridge (Engl 2002). Subsequent development at the neighbouring Dunure Road development revealed at least three post-built circular Mesolithic structures within a large multi-period site. The houses were typically sealed by an organic soil, with the right post cluster 4-5m in diameter with openings to the southeast. A lithic assemblage and coarse stone artefacts were also recovered while numerous pits, some with burnt organic matter, were located in proximity (Engl 2015). These discoveries have not reached publication as yet. 
  1. Road construction through Dunragit, near Luce Bay, revealed several discrete sites along the route, with Mesolithic activity very much in evidence. Archaeological investigations at West Challoch (7000 to 6000BC) recovered nearly 17000 lithics, in many instances associated with evidence for in situ working, as well as hearths (both stone settings and latent hearths) and three possible structures, one formed of two arcs of postholes interpreted as a substantial structure. The lithic analyses carried out by on the assemblage from West Challoch is the largest in the region and will provide a comparator for any late Mesolithic narrow blade assemblage (Ballin 2021). The researcher argues that the closest affinities are between the lithic industries of southern and eastern Scotland, contrasting with western Scotland. Residual evidence for early Mesolithic and Upper Palaeolithic activity was also found on this site. 
  1. Despite other regions becoming increasingly prominent, academic synthesis continued to see relevance from research in South West Scotland for wider discussions of Mesolithic Scotland. In 2004 the landmark conference on Mesolithic Scotland and its Neighbours (Saville (ed) 2004) the region was strongly referenced in both the author’s introductory paper. A subsequent paper analysing structural evidence from the Mesolithic (Wickham-Jones 2004) included Littlehill Bridge along with Barsalloch, Low Clone, Starr and Redkirk Point. However, the early 21st century saw a decline in the role of our region with the important synthesis of Mesolithic Scotland by Warren (Warren 2005) having none of its key sites in our region. Although Redkirk Point and Barsalloch were mentioned in the text, when discussing the role of fires and hearths for communities (Warren 2005, 125-127). 
  1. Until very recently, no robust evidence was present for Upper Palaeolithic occupation of the region – though broad blade Early Mesolithic lithic assemblages from Shewalton and Auchrocher Moss had been highlighted as potentially incorporating earlier material (Morrison & Bonsall 1989). Recognising Upper Palaeolithic activity can be achieved through contrasts between an earlier non-geometric blade industry and a later geometric blade industry. This means of differentiation proved crucial: it marked a division between the closing stages of the Upper Palaeolithic and the onset of the Mesolithic.  
  1. Since this distinction was made, evidence for this earlier technique has been identified amongst later, Mesolithic assemblages, indicating that there was a human presence in the region during the Upper Palaeolithic. Examples of these residual artefacts from South West Scotland are very limited, but include a zinken piercing tool (Hamburgian 12,700 to 12,200BC) from West Challoch and an unstratified blade from Girvan (Suddaby 2022) that may be either Late Upper Palaeolithic or Early Mesolithic. 

Leave a Reply