- Evidence of Neolithic occupation has been noted in several locations across South West Scotland, but actual traces of settlement are scarce. Neolithic sites are often identified through dense concentrations of worked stone tools and pottery sherds. Sometimes these are accompanied by clusters of pits and postholes, which suggest that structures may have been present. In most instances, however, these negative features do not appear to be arranged in any logical or coherent manner, which might denote the presence of a structure or structures.
- There are some exceptions. Developer-led archaeology has led to the discovery of several Neolithic timber halls, a group consisting of large rectangular timber structures dating to the earlier Neolithic, that is, during the period 4000 to 3400 BC. The interpretation of these structures as domestic houses has been questioned, as has the possibility that they were ever intended to be roofed. Several of these structures – including the example from Lockerbie Academy – were burnt down after use (Millican 2016), and in some cases, the quantity of burnt material present was sufficient to suggest a roofed structure. Their size, coupled with their relative scarcity, has led some to argue that they held a largely ceremonial role, perhaps serving several communities, as opposed to providing day-to-day accommodation for an extended family or small group. This extraordinary usage has, however, also been questioned: the excavators of one such structure at Laigh Newton argued that the focus of their particular investigations represented the ‘practical reality of life during the Neolithic rather than the ritual perception of life’ (Toolis 2011, 44). Possible timber halls have also been identified at Station Brae, near Irvine (Addyman 2004) and Hillhouse Farm, Kilmarnock.
- Occasionally, groups of postholes form potential structures which could feasibly be put forward as rare examples of Neolithic settlement. To date, all of these appear to be associated with Late Neolithic occupation. One example recorded as part of developer-led work at Doonhill, Arran, revealed several small, sub-oval post-built structures found in association with Late Neolithic Grooved Ware pottery (Gordon 2016). Similar small structures associated with Grooved Ware were reported at Station Brae (Addyman 2004). Another possible settlement site has been identified at Drummullan, although in this instance, the excavators, though acknowledging the potential for settlement, were cautious to ascribe the term (Wood 2020). The site of Beckton Farm, Lockerbie, revealed a circular ring-groove structure, again associated with Grooved Ware pottery (Pollard 1997). Here, the excavator also recovered evidence for a further two possible structures nearby, which were more ephemeral in character; one of these produced an earlier Neolithic date.
- More common are pit clusters associated with finds of Neolithic material culture and burnt material potentially derived from hearths or cooking pits, but which do not have associated discernible structures. Sites which fall within this group include Monkton, which was excavated in 2011. This site contained pits, postholes and a possible fire-pit, associated with Neolithic material culture and burnt hazelnut shells, but showed no clear focus or groupings, which might indicate structures. At Drummullan, 14 pits were identified within a similar pit cluster; these produced finds of Early Neolithic carinated bowl and radiocarbon dates between 3517 and 3368 BC (Wood 2020). Investigations at Ayr Academy revealed Middle Neolithic firepits, associated with quantities of charred hazelnut shell, with no associated evidence of cereal grain. Excavations at Barassie from 2008-12 revealed a similar occurrence of Neolithic pits and artefacts, but no apparent patterns that could denote the form and layout of actual structures. So did the site at Hilltop, near Twynholm, which revealed 10 pits which yielded sherds of carinated bowl and dates which spanned a range from 3640 BC to 3500 BC.
- Other sites have revealed evidence in the form of material culture, but are either largely or entirely lacking the accompanying pits and postholes. The coastal sites of Torrs Warren at Luce Bay and Shewalton at Irvine Bay are perhaps the best known of these sites, having produced vast quantities of artefacts, few of which are properly contextualised. Both sites are subject to constant natural reworking through dune movement and erosion, so even re-excavation, in the case of Torrs Warren, failed to reveal any firm evidence for settlement (Cowie 1996). A similar situation can be envisaged for Neolithic settlement sites located further inland, although in this case, the ongoing and regular disturbance is down to later human activities. It seems entirely likely that Neolithic settlement favoured sites which also proved attractive for later agriculture and occupation. Over millennia, this has led to the removal of ephemeral traces and the survival only of the largest, most robust features, which tend to be associated with ritual, as opposed to domestic, practice.