Current analysis and understanding of the Scottish Neolithic is narrated on a number of different chronological and spatial scales, ranging from broad-brush ‘grand narratives’ (as offered in Theme 2), through regional accounts, to landscape-, site- and artefact-level studies. How the articulation of these diverse scales of investigation and narrative-building can be improved is a key challenge.
For present purposes the focus shall be on developing narratives at a regional level, since this approach provides a sufficient spatial resolution to identify broad patterns, while taking cognisance of levels of detail which get lost in the national picture. The regional level also allows examination of what happened, when, at a greater level of detail than can be addressed through higher-level approaches.
3.2.1 Spatial resolution
As the ‘Big picture’ narrative of Theme 2 has made clear, following the initial appearance of two separate ‘strands’ of Neolithisation during the late 5th to early 4th millennium – the one (the Atlantic façade, ‘Breton’ Neolithic) appearing on parts of the west coast of Scotland, the other (the ‘Carinated Bowl’ Neolithic) more widespread but essentially south of the Great Glen, except for parts of the Caithness coast – there followed a process of regionalisation, as traditions became reproduced and reinterpreted over time.
The fact that the Scottish Neolithic was regionally variable has indeed long been recognised, although the way in which Scotland has been divided into regions has varied. Piggott’s (1954, 381) regional division of Scotland was essentially east vs. west, following his vision of two-strand Neolithisation and acknowledging that the Scottish Neolithic was part of a broader cultural tradition in Britain. More recent attempts to consider the regional nature of the Scottish Neolithic have focused on smaller areas, such as Orkney (e.g. A. Ritchie 2000), the Western Isles (Armit 2003) and SW Scotland (Thomas 2000), with Gordon Barclay arguing for the east Scottish lowlands (and particularly the lowlands of Aberdeenshire) as having a strong regional identity (Barclay 2000 – although note that some of the evidence for this distinctiveness – i.e. the recumbent stone circles and (two-entrance) henges – have since been shown to belong to the Chalcolithic and Early Bronze Age).
The definition of regionality between the late 5th and mid-3rd millennium BC is, however, problematic. As Niall Sharples noted (Sharples 1992), Scotland is not amenable to being divided into discrete Neolithic regions, since some traditions, monument types and styles of material culture transcend ‘natural’ geographical regions (as seen, for example, in the distribution of cursus monuments and bank barrows) and since changing patterns of interaction altered the distribution of shared cultural traits over time. Furthermore, the concept of what defines a region is a slippery thing. While the topography of Scotland suggests certain natural regional divisions and an overall ‘highland vs. lowland’ differentiation, and while topography will have influenced routes and ease of travel and communication, it clearly did not determine cultural identity. One could, of course, attempt to define regionally-variable archaeological cultures, using David L. Clarke’s concept of culture as ‘a polythetic set of specific and comprehensive artefact-types [and, by extension, site-types and traditions] which consistently recur together in assemblages within a limited geographic area’ (D.L. Clarke 1968, 232, with addition). However, the fierce debates of the 1970s and 1980s in British archaeology over the use (and utility) of the concept of archaeological ‘cultures’ highlighted the danger of such an approach, pointing out that a shared set of cultural attributes need not correlate with ethnic, linguistic or any other kind of cultural identity. Thus, in contrast to any Continental counterparts, no attempt will be made here to define ‘cultures’ as such (beyond the broadest level of ‘the Carinated Bowl Neolithic’, for example).
Other issues lurk behind the use of regionality as a concept. Firstly, as Gordon Barclay has pointed out (2000), by focusing archaeological endeavour and resources on a few regions – with Orkney being the Scottish example he cites – this skews any knowledge of other parts of Scotland, so that we end up knowing a great deal about a few areas, and relatively little about others. To some extent, however, this issue has been mitigated by the strategic use of resources by Historic Scotland over the last 15 years, and by the impact of developer-funded archaeology (which has tended to follow the geography of infrastructural and commercial development). And in any case, issues of how different regions are investigated is incidental to the characterisation of regionality.
Secondly, the current perspective may be skewed by a contemporary sense of regionality, which can often be defined by modern – and not-so-modern – administrative boundaries (e.g. local authority areas and pre-1975 counties).
Thirdly, there needs to be an acknowledgement that the Neolithic perception of space and identity may not have included any concept of regionality at all – beyond an awareness that other people lived in other areas, some a long way away. Groups will indeed have had an identity, and this will have had a geographical aspect and there may indeed have been some sense of territorial ownership (at a local level); but what is seen, when distribution maps of site and artefact types are examined, is the outcome of shared traditions, patterns of interaction and cultural choice.
Notwithstanding these considerations, the use of a regional approach does have heuristic value, for the reason given above – i.e. it allows patterns of behaviour to be investigated at a more detailed level than that of the ‘Big Picture’. Furthermore, there clearly was regional variability during the Neolithic, even though the geographical characteristics of this variability will have changed over the centuries; this variability needs to be characterised.
A pragmatic approach has therefore been followed, whereby regionality is used as an analytical and narrative device. Scotland (and indeed neighbouring parts of England) has been divided into broad geographical regions for the purpose of describing material culture, monuments, traditions and practices, on the understanding that:
- these regions are a largely arbitrary division, albeit one in which shared intra-regional, and divergent inter-regional patterns can be perceived; and
- there will have been inter-regional interaction, differing in its extent and direction at different times during ‘the Neolithic’.
After much discussion within the Panel, the regional divisions were delineated as follows (with relevant section numbers included)
3.3.1 East and central Scotland, between the Great Glen and the Forth
3.3.2 South-east Scotland and north-east England
3.3.3 South-west Scotland
3.3.4 Western Scotland, the islands of the Clyde and the Inner Hebrides, south of the Great Glen
3.3.5 The north-west mainland and the Hebrides north of the Great Glen
3.3.6 The north-east mainland (north of the Great Glen) and Orkney
3.3.7 Shetland
3.2.2 Chronological Resolution
Issues of chronological resolution can be considered from several perspectives: What level of chronological resolution is required (or desired) to achieve robust narratives at different scales? What are the limitations to the level of chronological resolution which can actually be achieved? What are the potential consequences of these limitations to narratives?
Traditionally, many higher level narratives for the Scottish Neolithic have been related to a level of chronological resolution provided by a period-based system, in which ‘the Neolithic’ is divided, either into two (i.e. Early vs. Late) or three (Early-Middle-Late); the arbitrariness inherent in any periodisation has already been touched upon in Theme 2.
The problems with a period-based system have been widely recognised, yet such systems still persist in use. While these may, arguably, act as convenient short hand (shorter still would be N1, N2 etc), the question remains: what defines these periods? In most cases, the definition / extent of a sub-period derives from recognition of the changing character of a range of archaeological evidence, frequently interpreted as reflecting or relating to other social or economic trends. Thus, the nature of any narrative will in part be defined by the degree of chronological resolution available.
Nowadays, with a growing body of reliable radiocarbon dates and the application of Baeysian modelling, the prospect of being able to identify developments over individual generations, rather than over centuries or quarter millennia (as had been the limitation in the past) is opening up (cf. Whittle et al. 2011). This means that finer-grained, more detailed narratives of what was happening, when, and to assess the speed of change will eventually be able to be created. Of course, the vagaries of the radiocarbon calibration curve mean that there will always be some periods when chronological resolution is poorer than others: within our time frame of interest, the plateau at 3500-3100 BC is particularly frustrating (cf. Schulting et al. 2010). Unless a significant amount of dendrochronologically-datable material is found, or ways of mitigating the problem of the calibration curve are discovered, then this will remain a challenge.
In the regional reviews that follow, a pragmatic solution has been chosen that enables readers to compare developments in different regions. Fairly arbitrary divisions have been made, as follows:
- From beginnings to 3500 BC (which will go by the shorthand term ‘Early Neolithic’)
- 3500-3000 BC (‘Middle Neolithic’)
- 3000-2500 BC (‘Late Neolithic’).
The reason why this does not exactly follow the chronological divisions used in Theme 2 is that the separate sections on ‘Beginnings’ and ‘Endings’ were used there to highlight specific issues for the overall narrative. Having dealt with those, what happened between the ‘Beginnings’ and the ‘Endings’ can be focused on.