Evidence for houses is very sparse indeed and, as noted above, there are no definite examples of the huge Early Neolithic ‘halls’ – communal houses for initial groups of farmers settling in an area – in Highland Region. (The nearest definite examples of ‘halls’ are those at Warren Field, Crathes and at Balbridie, both Aberdeenshire: Murray et al 2009; Ralston 1982). Likewise, no slot trenches for any plank-built Neolithic structures have been found, although these could easily have been destroyed by recent and historical ploughing.
The heavily truncated remains of a rectangular post-built structure, interpreted as a house, was discovered at Kinbeachie on the Black Isle (Fig.5.2; MHG58909; Barclay et al 2001). It measured 7 x 4 m, and the fills of the post-holes were relatively artefact-rich. The highest concentration of cereal grain within the excavated area was also located in this structure and this, together with the artefactual evidence, strengthens the suggestion that this was a house. The associated pottery is Impressed Ware, and associated dates suggest that this was in use during the second half of the fourth millennium.
A sub-rectangular post-built structure that appears to be a Late Neolithic house with a central hearth was found at Stoneyfield, Raigmore, on the outskirts of Inverness (Fig. 5.5; MHG45834; Simpson 1996a). Measuring 14 x 6 m, it pre-dated a Clava cairn that was constructed over the house during the late third millennium, and it appears to be contemporary with several pits containing Grooved Ware, including one (Pit 20) that contained cremated human bone that has recently been dated to 3090‒2907 cal BC (Copper et al 2018, 223. (Note that the cremated human bone from Pit 50 – another pit containing Grooved Ware, and with a chisel-shaped arrowhead – was found to be intrusive in the pit fill, relating to much later, Late Bronze Age secondary activity at the site: ibid 224.)
Derek Simpson argued that at least two pits (36 and 49, the latter containing Grooved Ware) pre-dated the structure. He commented that it was hard to resolve a coherent plan for the house from the pattern of post-holes (despite a clear sub-rectangular overall plan) and, given the presence of cremated human remains in some pits, he wondered whether this could have been a structure associated with funerary practices rather than a house; but the disposition of the post-holes does not resemble that of the (probably slightly earlier) putative exposure platforms in Tayside (eg at Balfarg Riding School, Fife: Barclay and Russell-White 1993). It seems more likely that the Raigmore ground plan reflects the building and rebuilding of a roofed structure. Dates later in the third millennium BC had previously been obtained by Simpson for charcoal from other pits containing Grooved Ware pottery (namely Pits 21, 49 and 19: see Datasheet 2.1 and Copper 2019 for details), but the charcoal in question could be intrusive from the time when the Clava cairn was constructed. (See also Copper et al 2021 for a discussion of the dating of Raigmore.)
A further possible structure of possible Neolithic date has been identified at Bellfield Farm, North Kessock (MHG53530; Jones 2009, illus 2 and 3). This consists of an arc of postholes around 5 m across, thought by the excavator to be either a small sub-circular building or else the rounded end of a sub-rectangular building. A cluster of pits nearby contained pottery identified by Julie Lochrie as Impressed Ware.
At Beinn Tighe on Canna, a low settlement mound (MHG5560) – one of five mounds on Canna known or suspected to be of Neolithic date – produced evidence for two stone structures, one of which is oval and measures 2.8 m by 2.2 m (Gannon 2016, 144). This, and possibly also the other structure, could well have been a house.
A clue to one constructional technique used for building Neolithic houses is offered by three fragments of daub, with wattle impressions, found at Fortrose and Rosemarkie Waste Water Works (EHG4280; Sheridan 2014a, 18). The daub would have served to render the wattlework walling wind- and waterproof. Sheridan noted that one piece of daub was found to have a fragment of a pitchstone blade accidentally embedded in it. Daub, or burnt clay, was also found at Culduthel Farm (Phases 7 and 8) in a spread of material believed to be of Neolithic date (Murray 2008).
It remains to be seen whether the recent developer-funded excavations in and around Inverness will produce any evidence for houses; in most cases it appears that settlement evidence consists of pits and/or artefact spreads.
Finally, it should be noted that there is no evidence for the use of roundhouses during the Neolithic period in Highland Region (or elsewhere in Scotland). At Auchtercairn, near Gairloch (MHG59292), for example, the dated Late Neolithic activity relates to the use of a hearth that is unrelated to the Bronze Age roundhouse in the vicinity (www.wedigs.co.uk/c14results.htm).