As with 'Argyll', the 'Palaeolithic' and 'Mesolithic' are themselves cultural constructs with no meaning for the early prehistoric hunter-gatherers themselves. The Palaeolithic is simply correlated with the Pleistocene, potentially extending back to pre-modern human presence, although the only existing evidence is restricted to the Late Glacial period. The Mesolithic starts with an environmental event, the Pleistocene/Holocene transition at 11,650 cal BP, and ends some time after the first appearance of a novel, Neolithic lifestyle based on agro-pastoralism, most likely with new communities arriving from the Continent. Funerary monuments and material culture that has stylistic similarities to the Breton Middle Neolithic can be situated at some time between 6300 BP and 5900 BP (Sheridan 2010). On the basis of a small sample of skeletal material from Oronsay, some people continued following a wholly Mesolithic lifestyle up to c 5800 BP (Charlton et al. 2016), but the extent to which this is reflective of the region as a whole remains unclear. While the Mesolithic is characterised as a period of hunting, fishing and gathering, it is likely that various forms of environmental manipulation and potentially management would have been undertaken.
Although definitions are relatively straightforward, identifying sites of the Palaeolithic and Mesolithic is more challenging because stone artefacts, which form the bulk of the surviving evidence, are not directly informative of either lifestyle or date. Leaving the Palaeolithic aside, Mesolithic sites within Argyll can be defined as those that 'possess either microlithic technology, one with prepared platform cores, blades and microliths, or an 'Obanian culture' with bevelled-ended artefacts associated with bipolar core technology and shell middens' (Wicks and Mithen 2014, 246). A range of coarse stone, antler and bone artefacts are also found at Mesolithic sites, as reviewed by Saville (2004b), Figure 26.
Figure 26: Typical artefacts interpreted as designating Mesolithic activity © copyright
The microlithic technology is generally referred to as the narrow blade tradition, a term used to distinguish this from the use of broad blade microliths as found in the Early Mesolithic in England. Broad blade microliths are known within Scotland and are prominent within some assemblages in Argyll, notably at Glenbatrick CANMORE ID 38261) and Lussa Bay (CANMORE ID 38650) on Jura (Saville 2004b). It has not, however, been possible to establish whether these constitute a chronologically early phase of the Mesolithic, as in England. Moreover, narrow blade assemblages in Scotland have now been dated to c. 10,400 cal BP from Cramond, in eastern Scotland, with their earliest unambiguous appearance in Argyll at c. 9000 cal BP at sites such as Rubha Port an t-Seilich (CANMORE ID 98306) on Islay and Fiskary on Coll (Mithen et al. 2015). At the other end of the Mesolithic time period, the Obanian was once proposed as a distinct late Mesolithic cultural phase. This is now recognised as a behavioural variant within the Mesolithic, with its supposedly characteristic shell middens, bevel ended artefacts and bipolar technology found throughout this period. Pirie has stressed technological continuity between the Obanian chipped stone tool assemblages as found on Oronsay and the narrow blade assemblages as found on the neighbouring island of Islay (Pirie et al. 2006; in Wicks et al. 2014).
Identifying archaeological sites as Mesolithic on the basis of technology as defined above is unsatisfactory for two principle reasons. First, it does not preclude the possibility that such technology continued into later prehistory, as is known to be the case with bevel-ended artefacts. While the Neolithic and Bronze Age periods are characterised as having farming lifestyles, it may well be that hunting and gathering continued as a supplementary means for securing not only food but a wide range of raw materials, and may have at times have become critical. As such, it is theoretically possible that stone technology similar to that used in the Mesolithic might have been adopted, such technology presumably having a high level of functionally efficient design for hunting and gathering activities. If so, narrow blade assemblages cannot necessarily be assumed to Mesolithic when that is defined as the period prior to the appearance of Neolithic culture. At Bolsay (CANMORE ID 37342 AND CANMORE ID 37347) on the Isle of Islay, for instance, an extensive scatter of artefacts dominated by narrow-blade technology provided a larger number of radiocarbon dates falling after the supposed transition to the Neolithic (c. 6000 cal BP) than those within the Mesolithic period (Mithen, Lake and Finlay 2000). However, whether the radiocarbon-dated material definitely dated the narrow-blade artefacts, or whether we are dealing with a palimpsest of Mesolithic and much later Neolithic activities, remains unclear.
A second matter for concern is the converse: stone technology without any evident narrow blade, broad blade or Obanian characteristics might have been used in the Mesolithic period. This is especially the case for localities in which stone without the classic conchoidal fracture of flint and chert might have been selected. Such stone might have required an alternative means of flaking or simply left minimal and ambiguous traces of its fracture patterns, as with quartz. As such artefacts made in the Mesolithic period might be designated to the Palaeolithic, Neolithic and Bronze Age, or simply not recognised as cultural artefacts at all.
In light of these two issues, we can only confidently attribute those sites with radiocarbon dates between 11,650 and 5800 cal BP as being Mesolithic, of which there are 32, distributed as in Figure 24. Three-quarters (24) of those sites are located in Argyll. This is, of course, only a small proportion of the number of probable (i.e. undated) Mesolithic sites that are known to exist, as defined by the artefact criteria specified above. Whether this 'small proportion' is closer to 50%, 20%, 5% or even 1% is unknown at present.
Within archaeology the term 'site' conventionally refers to a specific locality that had a continuous period of activity within a designated period. This is unsatisfactory when dealing with mobile hunter-gatherers because any one site might represent a palimpsest of debris arising from multiple short-term visits spread over a long period of time. As such the entire chronological range between the first and final visit is of limited value, because the hunter-gatherers might have been absent for the majority of that time. A more effective measure is to use the radiocarbon dates to assess the minimum number of activity (occupation) events required to have created that set of radiocarbon dates by clustering the statistical consistent dates as minimally indicative of a single event.
As an example consider the site of Fiskary Bay (CANMORE ID 299865) on the Isle of Coll (Mithen et al. 2007a; Wicks and Mithen, forthcoming), Figure 27. This has six AMS radiocarbon dates derived from single fragments of hazelnut shell excavated from within a palimpsest of microlithic artefacts, fish bones and charred plant remains. The dates range from 8200 ± 50 BP (c. 9110 cal BP) to 7460 ± 50 BP (c. 8300 call BP), which from a site perspective provides an occupation period of c. 800 years. The six dates, however, fall into three pairs of statistically consistent dates, indicating three discrete activity events - each of which could have lasted for no more than a few hours, Figure 28. The earliest is centred on 9070 cal BP, a second at 8520 cal BP and a third at 8280 cal BP. In this regard, rather than referring to Fiskary Bay as a single Mesolithic site, it is more accurate to refer to it as having a minimum of three separate Mesolithic activity events.
Figure 27: Fiskary Bay, a Mesolithic site on the Isle of Coll © copyright
Figure 28: Radiocarbon dates from Fiskary indicating three discrete activity events © copyright
A further important factor regarding the nature of the early prehistoric record is that the distribution of sites, and hence dated activity events, is inevitably strongly influenced by patterns of recent human activity that may be quite different to the distribution of Palaeolithic and Mesolithic activity. Much of Argyll is covered by peat and blown sand accumulated after the Mesolithic period. Only where this has either not accumulated or been cleared (by human means or erosion), does the possibility exist of finding early prehistoric sites. As such, their distribution is likely to be biased to areas of farming activity, urban development and so forth. As the recent discovery of Mesolithic artefacts in the Cairngorms have demonstrated (Fraser et al. 2013; Noble et al. 2014)), Mesolithic sites might exist in areas that are unlikely to be covered by intensive archaeological survey or where chance finds during development might occur.
The situation regarding site detection changes in the Neolithic due to the appearance of funerary monuments, which have introduced architectural permanence and high visibility in the landscape. The initial influx of monument builders, represented by small passage and closed chamber tombs like Achnacreebeag, Figure 29, seem to have been low in numbers and many areas, including most islands, seem to have been unaffected. What appears to have been a second wave of monument building from c.5800 cal BP, imposes the tradition of Clyde tombs which are found more extensively in north-west Britain than Argyll alone (Figure 28). Early Neolithic settlement are elusive, leaving open the nature of the Neolithic lifestyles with regard to mobility, seasonality and economy (Thomas 1996 , Sheridan 2010). As such, western Scotland remains quite unlike eastern Scotland and Ireland, where large timber halls and rectangular timber houses, respectively, provide examples of Neolithic 'domestic' architecture (Fairweather and Ralston 1993, Barclay et al. 2002, Grogan 2004, Brophy 2007, Murray et al. 2009).
Figure 29: The small passage and closed chamber tomb of Achnacreebaeg © copyright
Some Mesolithic sites, such as Bolsay (CANMORE ID 37342 AND CANMORE ID 37347) (Mithen 2000) and Newton (CANMORE ID 37769) (McCullagh 1989) on Islay and Kinloch (CANMORE ID 22202) on Rum (Wickham Jones 1990) continued to be visited in the Neolithic. Shell middens at Cnoc Coig (CANMORE ID 37818) on Oronsay and Raschoille cave (CANMORE ID 22924) and Carding Mill Bay (CANMORE ID 22947) near Oban also straddle Mesolithic-Neolithic transition, although the nature of their use in the Neolithic in certain cases may be altogether different from that of the Mesolithic (Milner and Craig 2009). More often than not the 'Neolithic' levels can be difficult to distinguish in regard to the nature of the stone artefact assemblages (Armit and Finlayson 1992), this stressing the need for rigorous sampling and dating of the entire stratigraphic sequences on such sites.
Early Neolithic artefact types, such as pottery, polished axeheads, leaf-shaped arrowheads, and plano-convex knives are indicators of change because these are absent in the Mesolithic (Figure 30). They are also useful in demonstrating contacts between different regions, as is the case with the Irish porcellanite and continental jadeitite axes in Argyll for example (Sheridan and Pailler 2012). Few examples of carved stone balls found in Argyll point to contacts with the east of Scotland and the Northern Isles (Marshall and Taylor 1977). In contrast the presence of the type technologies traditionally seen as 'classic' Mesolithic, such as narrow blade technology can occur either residually may have continued in use with the Neolithic and later periods.
Figure 30: Diagnostic artefact types for the Neolithic in Argyll © copyright
Pottery is relatively rare other than within funerary monuments, but when it occurs it can distinguish between Mesolithic and Neolithic deposits. At Newton (CANMORE ID 37769) on Islay, for instance, the presence/absence of pottery was used to distinguish some pits as Neolithic and others as Mesolithic (McCullagh 1989). Similarly the presence of pottery was used as the main demarcation between the Mesolithic and the Neolithic levels at Kinloch (CANMORE ID 22202), Rum, (Armit and Finlayson 1992). At Rubha Port an t-Seilich (CANMORE ID 98306) on Islay, where there is both late Glacial and Mesolithic deposits with culturally diagnostic chipped stone artefacts and secure dating, a fragment of what is likely to be early Bronze Age pottery was nevertheless recovered during excavation indicating later prehistoric activity.
In summary, the Mesolithic record is dominated by scatters of chipped stone artefacts derived from a range of manufacturing techniques including platform-core and bipolar technology. Designating these as necessarily Mesolithic in a chronological sense is problematic because such technology might continue in to later prehistory, while other Mesolithic sites (in the chronological sense) might be neglected because of an absence of evidence for such manufacturing techniques. The Early Neolithic is denoted by a series of type artefacts, although the absence of these might not preclude the possibility of a site as being of Neolithic date. The distribution of sites will be biased by numerous factors, including the extent and nature of farming activity, urban development and archaeological research. In the Neolithic there is an additional bias towards highly visible mortuary monuments in comparison to a very few known settlement sites.