3.3.1 East and central Scotland, between the Great Glen and the Forth

This region includes the rich agricultural lowlands of Tayside and Fife, Aberdeenshire, Moray and the Black Isle - areas that are rich in Neolithic archaeology, but which have also seen millennia of farming, which has created a palimpsest of later activity and taken its toll on the Neolithic remains. Current knowledge of the Neolithic of this region owes much to research-based fieldwork, survey and synthesis, of which a considerable canon has grown up over the course of the 20th and early 21st century. Highlights include the following:

Research excavations and field survey:

  1. Early Neolithic timber 'halls' on either side of the River Dee at Crathes Warren Field (Murray et al. 2009) and Balbridie, Aberdeenshire (Ralston 1982; Fairweather & Ralston 1993), and at Claish, Stirling (Barclay et al. 2002);
  2. Non-megalithic round mounds at Boghead, Moray (Burl 1984), Pitnacree, Perth & Kinross (Coles and Simpson 1965) and Pitglassie, Aberdeenshire (A. Shepherd 1996);
  3. The Cleaven Dyke and the Littleour structure, Perth & Kinross (Barclay & Maxwell 1998)
  4. Stuart Piggott and Derek Simpson's excavations at Croft Moraig, Perth & Kinross (Piggott & Simpson 1971) and its subsequent reappraisal (Bradley & Sheridan 2005);
  5. Alan Saville's excavations of flint mines at the Den of Boddam and on Skelmuir Hill, near Peterhead, Aberdeenshire (Saville 2008; 2011);
  6. The SERF (Strathearn Environs & Royal Forteviot Project) excavations of the Middle Neolithic cemetery, Late Neolithic timber enclosure, timber circle and subsequent activity at Forteviot, Stirling (Noble & Brophy 2011a; 2011b); and
  7. Aerial reconnaissance by RCAHMS and Aberdeenshire Archaeology (e.g. Shepherd & Greig 1996).

In addition to these, Richard Bradley's excavations of Clava cairns (Bradley 2000), recumbent stone circles (Bradley 2005; and cf. Welfare 2011) and small henges (Bradley 2011) in this region have served to demonstrate that these monuments actually post-date the conventional end-date of the Neolithic, belonging instead to the Chalcolithic and Early Bronze Age.

Synthetic reviews:

  1. Ian Kinnes' review of non-megalithic long mounds (Kinnes 1992a; 1992b) and non-megalithic round mounds (Kinnes 1979);
  2. Andrew Dunwell and Ian Ralston's review of the archaeology of Angus (Dunwell & Ralston 2008);
  3. reviews of the region's Early and Early to Middle Neolithic pottery by Audrey Henshall (Henshall 1968; 1983) and Trevor Cowie (Cowie 1992; 1993)
  4. Gordon Barclay's reviews of the archaeology of lowland Scotland (Barclay 1992; 1995; 1996; 2000; 2003)

Developer-funded (and other 'rescue') excavation has also played a major role in revealing the wealth of Neolithic sites in this region (Phillips & Bradley 2004), as demonstrated for example in the excavations in the Balfarg and Balbirnie complex, Fife, led by Roger Mercer (Mercer 1981) and by Gordon Barclay and Chris Russell-White (Barclay & Russell-White 1983); see also Alex Gibson's recent dating of the Balbirnie stone circle (Gibson 2010). Other important Neolithic sites revealed through rescue excavation include the complex of sites around Kintore, Aberdeenshire (Alexander 2000; Cook & Dunbar 2008); others around Inverness, especially at Culduthel (Murray forthcoming) and Raigmore (Simpson 1996); and the important Middle Neolithic site at Meadowend Farm, Clackmannanshire (Jones 2006; Jones & Smith forthcoming).

It should be noted here that, as far as the archaeology of Aberdeenshire and Moray is concerned, a very great debt of gratitude is owed to the late Ian Shepherd, Aberdeenshire Regional Archaeologist, who did so much to encourage and facilitate much of the work mentioned above in this part of Scotland. Early Neolithic, to 3500 BC

Figure a: Distribution of non-megalithic long barrows in Scotland. Based on Kinnes 1992a, with addition; note that one mound (at Meadowfoot, Dumfries & Galloway) may be a natural glacial mound

The Early Neolithic of eastern and central Scotland is marked by the appearance and subsequent development of the Carinated Bowl (CB) Neolithic, as sketched in Theme 2. Aberdeenshire is particularly rich in CB Neolithic sites, and these tend to cluster along major rivers and in areas which would have been favourable to mixed arable-pastoral farming (especially around Inverurie). This is shown, for example, in the distribution of CB pottery (Figure 7), of Alpine axeheads (Figure j) and of non-megalithic long barrows (Figure a); cf. the distribution of non-megalithic round mounds in eastern Scotland, Figure b.

Figure b: Distribution of non-megalithic round barrows in Scotland. From Sheridan 2010b

Indeed, that Aberdeenshire (and Angus) may have been an important landing place for putative immigrant farmers from northern France is suggested by the large houses ('halls': Figure c) at Crathes and Balbridie, on either side of the River Dee (with a further probable example known from aerial photography at Noranbank, Angus: Brophy 2007). These, together with the Claish 'hall' in Stirling, are among the earliest evidence for any Neolithic presence in this region and, as argued elsewhere (e.g. Sheridan 2010a and in press), these structures could well have been the communal houses for the first few generations of settlers, until they became sufficiently well established to 'bud off' into smaller, individual family settlements (at which point the 'halls' seem to have been decommissioned by being burnt down). The latter are represented, for example, by the houses (Figure 8) at Garthdee Road, Aberdeen, Aberdeenshire (Murray & Murray 2005); Pitlethie Road, Leuchars, Fife (Cook 2007); and, in a more truncated form, Deer's Den, Kintore, Aberdeenshire (Alexander 2000, illus 5) and Coul Brae, Mosstodloch, Moray (Gray & Suddaby in press). Other Early Neolithic settlements are represented by pits and/or artefact spreads, as for example in the case of the pit with 'traditional CB' pottery at Hatton Farm, Elliot, Angus, dating to 4930±30 BP, SUERC-24912 and 24913, 3780-3650 cal BC (Gray & Suddaby 2010), and the pits at Blackhall Road, Inverurie, Aberdeenshire (Lochrie 2010a).

Figure c: The 'halls' at Balbridie (top) and Warren Field, Crathes (bottom), both Aberdeenshire. Reconstruction drawings by David Hogg and Jan Dunbar with Hilary & Charlie Murray. Reproduced by permission of Historic Scotland and Murray Archaeological Services

Bayesian modelling of the dates for the Crathes, Balbridie and Claish 'halls' has indicated that their currency spans between 3800-3705 cal BC and 3705-3630 cal BC (at 95% probability) (Whittle et al. 2011, 832-3 and fig. 14.173), and modelling of the dates for the Garthdee Road house (ibid., fig. 14.157) has produced a near-identical range, suggesting that the 'budding-off' process started within a few generations of the initial arrival of the farmers. Furthermore, the presence of Arran pitchstone at Crathes (Warren 2009) indicates that the occupants participated in a network of contacts over which this material travelled a long distance from its source; this confirms that the process of establishing contacts between different groups of farmers started very early - thereby establishing a viable 'breeding population', apart from providing inter-group support.

In addition to housing several families, the large 'halls' would have constituted a powerful statement of identity and presence and, as noted in Theme 2, they would have constituted a very striking novelty, totally unlike any structure previously built in Scotland. Their Continental ancestry has been discussed elsewhere (most recently in Sheridan in press). That these were not just places of habitation, but were central to the belief system of their occupants, is suggested by the ceremony attending the construction of the Crathes 'hall', where the building of the structure was preceded by the erection of two massive, non-structual timbers along what was to become the long axis of the building (Murray et al. 2009, 37-40). These echo the split-trunk timbers used to frame the mortuary structures described below, and an analogy with totem poles has been drawn; at the very least, they indicate that wood/trees played a prominent role in the expression of beliefs (cf. Noble 2006).

As regards the susbsistence strategy of these early farming groups, palaeoenvironmental work around the Crathes structure has established that cereals were being cultivated in at least one field in the immediate vicinity (Lancaster et al. 2009), indicating settled agricultural habitation of the area, and at Balbridie, a particularly detailed impression of the crops grown was possible, largely thanks to a deposit of 20,000 carbonised grains found inside the structure (Fairweather & Ralston 1993). Whether these represented a food-store to tide the occupants over the winter, or else constituted a votive deposit at the time of the house's destruction by fire, is unclear. The Balbridie evidence demonstrated the cultivation of the environmentally-demanding bread wheat, as well as emmer wheat, barley, oats and flax; all but flax were also found at Crathes (Lancaster et al. 2009). Both sites also produced evidence for the gathering of wild plants, including hazelnut and (at Balbridie) crab apple. While animal remains were poorly preserved at these (and other) large houses, lipid analysis of the Crathes pottery has revealed that the occupants' cattle (and/or sheep/goats) had been used for dairying, since traces of ruminant milk fats were found in them (as well as pig fat:Šoberl & Evershed 2009). The human remains that survive from Early Neolithic eastern and central Scotland are either too small to permit isotopic analysis for obtaining dietary information (in the case of Dalladies) or else are not in a condition to permit such analysis, being calcined. The fact that large numbers of Neolithic arrowheads and other artefacts have been found on the sandhills of Tentsmuir in Fife and Culbin Sands in Moray (Clarke D V 2004) does not necessarily indicate activities connected with exploiting marine resources. The small size of many of the leaf-shaped arrowheads from these sandhills sites could as easily indicate their use in hunting birds as in shooting fish.

Figure d: Plan of Pitnacree round barrow, Perth & Kinross, showing the phases of construction and detail of the mortuary structure. From Kinnes 1992b

The funerary practices of these first farming communities in this region featured the use of non-megalithic structures (Kinnes 1992a, 1992b). Rectangular 'linear zone' mortuary structures (Figure 10), of timber construction (rebuilt, at Dalladies and Pitnacree, using dry-stone walling), were probably used for the laying out (and possibly natural excarnation) of the dead. In addition to the two examples mentioned above, a mortuary structure is known from Fordhouse Barrow, Angus (Proudfoot 2008). After an interval - during which the latest versions of the mortuary structures and their contents were burnt - these were covered over by earthen mounds, either long and rectangular or trapezoidal (as at Dalladies: Piggott 1972), or round (as at Pitnacree: Figure d; Coles & Simpson 1965). Distributions of these monuments are shown in Figure a and Figure b.

Large, rectangular pit-defined and post-built enclosures (Figure e) may also have been used for the laying out (and natural excarnation?) of the dead, as mortuary enclosures; examples include Inchtuthil, Aberdeenshire (Barclay & Maxwell 1991); Cowie Road, Bannockburn, Stirling (Rideout 1997), Douglasmuir, Angus (Kendrick 1995 - and see her illus 8 for further examples) and Castle Menzies, Perth & Kinross (Halliday 2002). As will be seen below, however, there is a question mark as to whether these monuments formed part of the earliest Neolithic activity in the area.

Figure e: Rectangular timber-built mortuary enclosure at Douglasmuir, Angus. From Brophy 1998

In addition, the practice of cremation on an open pyre, followed by the sealing of the site by a round or long mound, is attested at several sites: those associated with round mounds are Boghead, Moray: Burl 1984; Midtown of Pitglassie, Aberdeenshire (Shepherd 1996); East Finnercy, Aberdeenshire (Sheridan 2010b) and the 'Pow Sod' cairn at Atherb, Aberdeenshire (ibid.), while those associated with long mounds are Knapperty Hillock (Kinnes 1992a, 85) and one of the Cairns of Atherb (Sheridan 2010b).

The use of long and round earthen mounds to cover pyres and mortuary structures indicates that these monuments were not intended for successive interment; rather, they would have stood as highly-visible monuments and as statements of identity (and possibly also land ownership) and would have endured as such long after the large 'halls' had been burnt down. Perhaps they were regarded as houses for the dead, and particularly of the founding ancestors. Indeed, with the exception of the Cairns of Atherb, the number of individuals whose remains have been found in these monuments is small; and at Dalladies, just one (unburnt) fragment of a child's skull was found. (Also present at Dalladies, but not associated with the bone, was a small slab with nine cupmarks, found over a posthole: Piggott 1972.Cf. a slab with two cupmarks from a putative Early Neolithic context at Raigmore, Highland: Simpson 1996, illus 9).

As for the dating of the long and round barrows, Bayesian modelling of the dates for the former (taking into account that many of the dates are from oak, with a possible old wood effect) has concluded that the earliest dated examples were built within the time range 3935-3750 cal BC (94% probability), probably 3840-3775 cal BC (68% probility), and the latest examples were built 3760-3620 cal BC (95% probability) (Whittle et al. 2011, 828-9, fig.14.167). This indicates that they were built at least as early as the 'halls' and early smaller houses. As for the dating of the round barrows, this is problematic, due to the shortage of reliable and informative dates (Sheridan 2010b). Whittle et al. have claimed (2011, 830-2) that they started to be built after the long barrows, but in fact the existing dating evidence does not allow this to be determined either way (although they could have been constructed over the course of several centuries). The 'traditional Carinated Bowl' pottery at Pitnacree should be as early as any other example of this earliest form of 'CB' pottery in Scotland; it is particularly regrettable that the cremated bone associated with the primary use of this monument cannot be found (Sheridan 2010b).

Within the region, there are a handful of megalithic monuments (Henshall 1963, map 7), but these are all at the south-west and north-west fringes of the area in question and are understandable in terms of developments elsewhere in Scotland: the passage tombs around Inverness can be related to the passage tombs to the north of the Great Glen, while the Clyde cairn and 'unclassified' monuments in the south west, on the edge of the central Highlands, are outliers of a development that took place in western Scotland (see below). Human bone from one of these, the 'Clyde cairn' at Cultoquhey (Perth & Kinross), has been radiocarbon dated as part of Rick Schulting's programme of dating Scottish chamber tombs and has produced a date of 4680±40 BP (GrA-26922, 3620-3370 cal BC: Schulting pers. comm.; see Schulting et al. forthcoming for a discussion). This indicates that the use of this tomb did not fall within the earliest Neolithic in eastern Scotland.

Figure f: Aerial photograph and plan of cursus at Milton of Guthrie, Angus. From Brophy 1998; Crown copyright

A claim has been made for the existence of a (non-megalithic but stone-built) passage tomb in the complex, multi-phase monument at Fordhouse Barrow, Angus (Proudfoot 1998), but the structure bears an unmistakable resemblance to a keyhole-shaped corn-drying kiln, and until the full final excavation report has been produced, it would be unwise to speculate further.>

Cursus monuments (Figure f) and bank 'barrows' (Figure g) (with the Auchenlaich example actually a cairn) need to be mentioned here, since at least some of these structures may have been in existence by 3500 BC, even though they do not belong to the earliest Neolithic.

As discussed by Brophy (1998, 1999, 2006), these may represent aggrandised versions of the aforementioned mortuary enclosures and long barrows respectively; furthermore, it has been argued that the monument at Auchenlaich may have started its life as a Clyde cairn, with its mound subsequently elongated to an enormous 342 metres (Foster & Stevenson 2002). As Fig. k shows, cursus monuments and bank barrows tend to cluster in Tayside (incidentally complementing the distribution of long barrows), but their distribution extends beyond east Scotland to include other areas of 'CB Neolithic' Scotland, with a cluster around Dumfries (Thomas 2007).

Figure g: Auchenlaich bank 'barrow' (actually cairn). From Brophy 1998; Crown copyright

The chronological relationship between cursus monuments and mortuary enclosures has been explored by Whittle et al. (2011, fig. 14.170), who have Bayesian-modelled the currently available dates for the cursus at Holm, Holywood North and Upper Largie (all outside the region under scrutiny here); and for mortuary enclosures at Douglasmuir, Castle Menzies, Inchtuthil and Cowie Road, Bannockburn. They have concluded that the latter might not have pre-dated the former by very long (if at all), and furthermore have argued that none of these monuments demonstrably pre-dates 3700 BC, seeing the most likely date 'in the middle rather than the early centuries of the fourth millennium cal BC' (ibid., 830). As for the best-explored bank barrow - the c. 2 km-long Cleaven Dyke (Barclay & Maxwell 1998) - the date for its construction remains a mystery. The possibility that it may end with (or abut) a round mound requires investigation.

Figure h: Unstan Bowl from Spurryhillock, Aberdeenshire.

The material culture associated with the Early Neolithic in eastern and central Scotland is explored in Theme 5; suffice it to note here that the process of 'style drift' from the 'traditional CB' canon of pottery manufacture seems to have started very soon after this tradition of pottery appeared, to judge for the near-identity of the dates for the Crathes and Balbridie 'halls'. (The former's ceramic assemblage is of 'traditional CB' pottery, the latter's, of 'modified CB pottery' in what Audrey Henshall has called the 'North-Eastern' style - henceforth 'CBNE'.) The trajectory of development includes some features that show clear continuity from the 'traditional CB' (such as the increased use of fingertip fluting, and of ripple burnishing, as a decorative surface finish: Fig. l), and others that show a more marked departure. The latter include the use of lugged vessels (Fig. m) and, at Balbridie, of two bipartite collared pots that appear to form the prototype of the 'Unstan Bowl' (Fig. n, top). Intermediate forms between these and 'classic' 'Unstan Bowls' (as seen, for example, at Spurryhillock, Aberdeenshire) are known from Culduthel, Highland (Figure h).

The small flaked lithic tools of Neolithic eastern Scotland have usefully been reviewed by Graeme Warren (Warren 2006), who concludes that they are consistent with early Neolithic assemblages elsewhere in Britain and Ireland, and are characterised by the use of a platform technology (with some bipolar technology use as well) and by the production of artefacts such as leaf-shaped arrowheads (Figure 97), serrated blades and flakes, convex scrapers and plano-convex knives. This industry stands in contrast with Mesolithic flaked stone industries.

Figure i: Jadeitite axehead from Garvock, Aberdeenshire (Aarhus University, Department of Photo and Media, Rógvi N. Johansen). © Nationalmuseet

Other forms of Early Neolithic material culture in eastern and central Scotland include fine, non-utilitarian axeheads of Alpine jadeitite (and other Alpine rock), of which a cluster of examples are known from Aberdeenshire (Figure i).

Figure j: Distribution of Alpine axeheads in Scotland and the rest of Britain, Ireland and the Channel Islands. Image: Projet JADE

By analogy with the Alpine axeheads from Cairnholy and from the Sweet Track in the Somerset Levels, these axeheads are likely to have been brought over by the immigrant farmers, as ancient and treasured possessions, and to have been deposited shortly thereafter (Sheridan et al. 2010). One fragment of such an axehead, from Inverness (Fig. r), had been deliberately broken and burnt prior to its deposition close to the river.

Whether the equally fine all-over-polished flint axeheads (Figure k) also belong to the pre-3500 period is unknown, since none has been found in a datable context.

Similarly, it is uncertain whether the axeheads of Irish porcellanite that have been found in eastern Scotland (Figs. 41, and Figure l) arrived prior to 3500 BC; it is possible, since the CBNE style of pottery shares features in common with that belonging to the second quarter of the fourth millennium in north-east Ireland (and indeed west and south-west Scotland). The Great Glen appears to have been a route used in travel between north-east Scotland and west/ south-west Scotland and beyond, as illustrated by the close similarity between one particular kind of CBNE pot and its counterparts in west/south-west Scotland (Fig. 43). The example of this kind of pot from Culduthel, on the outskirts of Inverness, is from a context dated to c 3600-3500 BC.

Figure k: All-over-polished flint axehead from Bolshan Hill, Angus. ©NMS

Other external contacts, this time with Yorkshire, are suggested by the fine necklace of Whitby jet 'monster beads' and amber beads, found with a Yorkshire flint axehead, at Greenbrae near Peterhead (Figure 79; Kenworthy 1977; Sheridan & Davis 2002; and see below, see Section 5.2.5). The whole assemblage seems to have been imported from Yorkshire. A further 'monster bead', of locally-available material (probably a canneloid shale), has been found at Pitlethie Road, Leuchars, Fife (Sheridan 2007a). The dating of these beads is discussed in that publication and below; on currently-available evidence, a date within the second quarter of the fourth millennium seems likely. This jewellery suggests an interest in signalling status, and hence the existence of some degree of social differentiation, among the early farming communities of eastern and central Scotland.

Figure l: Distribution of Antrim porcellanite axeheads. From Sheridan 1986; note that many more are now known from Ireland Middle Neolithic, 3500-3000 BC

It is clear that the farming communities that had been established in eastern Scotland during the first quarter of the fourth millennium flourished and expanded during the second half of that millennium, since a number of settlements are known from that period. These include the following, to name but a few:

  1. Deer's Den, Kintore, Aberdeenshire (Alexander 2000)
  2. Midmill, Kintore, Aberdeenshire (Lochrie 2010b)
  3. Highfield, Inverurie, Aberdeenshire (Murray 2010)
  4. Alloa, Clackmannanshire (Mitchell et al. 2010)
  5. Balfarg Riding School (Barclay & Russell-White 1993)
  6. Grandtully, Perth & Kinross (Simpson & Coles 1990)
  7. Dubton Farm, Brechin, Angus (Cameron 2002)
  8. Newton Road, Carnoustie, Angus (White et al. 2009)
  9. Drumoig, Fife (James & Simpson 1997; Simpson 1997)
  10. Meadowend Farm (Upper Forth Crossing), Clackmannanshire (Jones & Smith forthcoming).

Further assemblages are listed by Trevor Cowie in his survey of the pre-Late Neolithic pottery of eastern and central Scotland (Cowie 1992; 1993) and yet more, like the Meadowend Farm and Drumoig material, await publication. The latter include assemblages from Culduthel, Highland and Powmyre Quarry, Glamis, Angus.

The nature of the habitation evidence was usefully reviewed by Gordon Barclay (Barclay 2003). In most cases the evidence consists of pits - often clustered - and artefact spreads, but there are hints of structural traces (e.g. at Grandtully). The heavy truncation of all these sites, through millennia of ploughing, will have erased all but the deepest of the features, but at Chapelfield, Cowie, Stirling, remains of circular stake-built houses were found (Barclay 2003, fig. 8.5) while at Kinbeachie (Highland) on the Black Isle - immediately outside the 'eastern Scotland' region as defined above but clearly related to east Scottish developments - traces of a rectangular house were found and radiocarbon dated to c. 3500-3100 BC (Barclay et al. 2001). Evidence relating to the subsistence strategy of the inhabitants is generally limited to traces of plant material (e.g. wheat and naked barley grains, fruit seeds and nutshells - e.g. Barclay & Russell-White 1993; White et al. 2009); bone preservation is generally very poor. As with the Early Neolithic material discussed above, it is clear that people were living in (or at least exploiting resources in) coastal areas, including sandhills - as attested, for example, by the assemblages of Impressed Ware from Brackmont Mill and Tentsmuir, Fife (Longworth et al. 1967).

One striking feature of many of these Middle Neolithic settlements is that they are in areas with earlier Neolithic (and indeed Later Neolithic) traces of settlement as well. This is clearly the case, for example, at Deer's Den, Dubton and Balfarg. (See Gibson 2010, illus 3, for a recent graphical representation of the spatial relationship of the different phases of activity.) In other words, there is evidence for continuity of settlement, even if the different episodes occurred in slightly different areas and were separated in time.

Evidence relating to Middle Neolithic funerary practices is very sparse and, with the exception of the aforementioned unburnt human remains from the Clyde cairn at Cultoquhey and the cremated bone found at Leven, Fife (see below), is limited to the large timber structures (Fig. 45) from Balfarg Riding School, Fife and elsewhere (Barclay & Russell-White 1993, 76-88; cf. Gibson 2010) that are thought to be mortuary structures. As noted by Barclay and others (including Brophy 2007), they resemble the Early Neolithic 'halls' in ground plan and size, although it has been argued that, unlike the 'halls', these Middle Neolithic structures seem not to have been roofed (Barclay 2003, 76). Their presumed use has been for the laying out of the deceased, to allow excarnation by natural processes (e.g. bird scavenging). In addition to the two Balfarg examples, further examples were excavated at Littleour and at Carsie Mains, Perth & Kinross (Barclay & Maxwell 1997; Brophy & Barclay 2004); part of another is suspected to have existed at Drumoig (Barclay 2003; Simpson 1997); and others are known from aerial photographs, in Tayside (Barclay 2003; Brophy 2007). As ever with regard to sites on gravel, unburnt bone will not have survived at these sites, although at Balfarg there are intriguing references to a deposit of calcined bone - presumed to be human, but sadly not identifiable now as mixed with another set of bone (Barclay & Russell-White 1993, 77). There were also finds of calcined sheep bone, and even - remarkably - traces of fish (ibid., 88).

The aforementioned cremated human bone from Holly Road, Leven, Fife is of particular interest since it consisted of two deposits, one of which was found in a segmented, arc-shaped ditch and dated to 4480±60 BP (GrA-21729, 3370-2930 cal BC at 2σ: Lanting 2004). Contrary to the excavator's view that this ditch was an enclosure for an Early Bronze Age cist cemetery (Lewis & Terry 2004), it appears that it was in fact a Middle Neolithic Age feature - hence the presence of the bone deposit, in a pit that had been cut into the ditch. A second deposit of cremated bone was found - as residual material - in one of the Early Bronze Age cists. This produced an earlier radiocarbon date (albeit one with a larger standard deviation) of 4760±90 BP (GrA-21728, 3710-3360 cal BC at 2σ), suggesting some time depth to the Neolithic deposition of cremated bone.

As for other kinds of monument, as noted above, it is unclear whether the enormous bank barrow-cum-cursus, the Cleaven Dyke, was constructed during the second half of the fourth millennium; the possibility that the construction and use of other cursus monuments and bank barrows/cairn extended into this period can also not be ruled out.

Middle Neolithic material culture is well represented among the numerous pottery assemblages, and a trajectory of pottery development that moves progressively further away from the traditional CB pottery of the Early Neolithic can be traced. The 'modified CB' pottery of the Middle Neolithic (Fig. 46) is generally thicker and significantly coarser than the 'traditional CB' repertoire, and the 'style drift' noted above included a decreasing use of carinated bowls and the use of other, novel vessel forms. This process can be seen, for example, at Balfarg Riding School (in comparing Cowie's 'Group 1' and 'Group 2' assemblages: Cowie 1993) The use of pottery decorated with various forms of impression (and sometimes also incision) - known by the general term of 'Impressed Ware' - can be regarded as part of the overall 'style drift'. The Impressed Ware assemblage from Dubton Farm, Brechin, Angus (Fig. 47), appears to represent an early version of this style of pottery and is associated, in one pit, with a date of 4735±40 BP (GU-9094/AA-39948, 3630-3380 cal BC at 2σ: Cameron 2002; MacSween 2007). A similar (and probably similarly early) assemblage is known from the aforementioned settlement at Kinbeachie, Highland (MacSween 2001). From around 3300 BC (or, at least, at some time between 3350 and 3000 BC, a variety of distinctively-shaped vessels with narrow flat bases came into use as a major element in the ceramic repertoire (Figure m).

Figure m: Example of truncoconic Impressed Ware bowl from Meadowend Farm (Upper Forth Crossing), Clackmannanshire. Reproduced courtesy of Headland Archaeology Ltd

Some of these were deep jars of various forms, others squat bowls, and most have a bipartite profile with a shallow upright neck. Rims can be heavy and club-like. Some vessels are undecorated, while others have profuse decoration on the rim and neck. Assemblages including this kind of 'Impressed Ware' include Deer's Den, Kintore, Aberdeenshire (Alexander 2000, illus 26), Brackmont Mill and Tentsmuir, Fife (Longworth 1968) and Meadowend Farm, Clackmannanshire (Sheridan 2011). This style of pottery was also used further south in Scotland (where it has been termed 'Meldon Bridge style' Impressed Ware: MacSween 2007) and in northern England; indeed, some of the large jars (e.g. Alexander 2000, illus 26, P49) resemble the 'Fengate ware' 'variant' of 'Peterborough Ware' in England.

As for other aspects of material culture, the use of leaf-shaped arrowheads and other Early Neolithic styles of small lithic tool seems to have continued (to judge, for instance, from the Kinbeachie assemblage: Wickham-Jones 2001). Similarly, pitchstone from Arran continued to be imported to eastern Scotland (e.g. at East Lochside, Angus: Ballin 2009,37), and links with Yorkshire continued - as demonstrated by the distinctively-shaped 'Duggleby adze' of flint found at Kemback, Fife (Fig. 49). Such objects are known to have been made in Yorkshire and recent dating of the eponymous find, from a grave in a round mound (Gibson 2009), has shown that they belong to the last quarter of the fourth millennium BC. Further evidence for links with Yorkshire during this period comes from Torben Ballin's recent analysis of flint assemblages: for example, he has observed that a significant proportion of the flint from Middle Neolithic pits at Midmill, Aberdeenshire, is likely to have been imported from Yorkshire (Ballin 2011, 58-60), while at East Lochside in Angus, a smaller proportion of the lithic assemblage is of grey Yorkshire flint (Ballin 2009, 37).

Figure n: Discoidal flint knife, Pitforthie, Aberdeenshire.© NMS

It is a moot point as to whether the (virtually) all-over-polished flint knife from Pitforthie, Aberdeenshire (formerly Kincardineshire; Figure n) constitutes a further Middle Neolithic import from Yorkshire. A Yorkshire provenance seems likely given that it is known that such objects were being made in North Yorkshire by specialist flintworkers (Durden 1995). A finer, thinner version of this artefact type was found in Burial D under the round mound at Duggleby Howe, Yorkshire, and the recent radiocarbon dating of this monument by Alex Gibson has concluded that the associated individual would have been interred shortly before the mound was raised at 2950-2775 cal BC (95% probability: Bayesian-modelled date, Gibson 2009, 69). Whether this means that this particular variant of so-called 'discoidal' knives (Clark 1932) was in use before 3000 BC, or should be regarded as a Late Neolithic artefact type, is open to discussion. (See below,, for more circular discoidal flint knives.)

Figure o: Distribution of axeheads of tuff from Great Langdale, Cumbria. From Clough & Cummins 1988

It is likely that at least some of the axeheads of Antrim porcellanite (Figure l), and the rather fewer axeheads of tuff from Great Langdale in Cumbria (Figure o) found in eastern Scotland arrived during this period. Indeed, a fragment of a Great Langdale axehead was found in a Middle Neolithic context at Midmill, Aberdeenshire (Ballin 2011, 59). In other words, the inhabitants of eastern and central Scotland continued to interact with communities elsewhere, over networks of contacts that had been established long before.

To this period also belongs evidence for flint mining on the Buchan Ridge Gravels near Peterhead (Figure p; Saville 2008; 2011); radiocarbon dating suggests that this activity probably occurred during the last quarter of the 4th millennium BC. The flint cobbles are too small to have been used for the manufacture of large items such as axeheads; evidence for the use of a Levallois reduction technique suggests that at least some of the flint was used for the production of blades for chisel-shaped arrowheads. Because flint of the same geological origin occurs elsewhere in this part of Scotland, however, it is not possible to document the distribution of the mined flint.

Figure p: Plan of the flint mines at Den of Boddam, Aberdeenshire. From Saville 2008; plan by Marion O'Neil Late Neolithic, c 3000-2500 BC

The widespread appearance of novel practices and novel material culture (including Grooved Ware pottery) around the beginning of the third millennium, as noted in Theme 2, is well attested in eastern Scotland. Finds of Grooved Ware have increased significantly since 1999, when only a dozen findspots from the whole of eastern Scotland were known (Longworth & MacSween 1999; cf. Cowie & MacSween 1999). Recent discoveries include the magnificent segment of a pot found in Powmyre Quarry, Glamis, Angus (Figure 98), along with large fragments from Midmill, Kintore, Aberdeenshire (Lochrie 2010b, illus 5), Inverurie Paper Mill, Aberdeenshire (Hilary Murray pers. comm.) and Mountcastle Quarry, Fife (inf Headland Archaeology Ltd). Such finds suggest a specific depositional practice - of placing large parts of individual vessels in pits, either singly or severally - that contrasts with the mode of deposition for earlier Neolithic pottery (whereby smaller sherds, and smaller portions of pots, generally tend to be found). In most cases - but by no means every case - Grooved Ware has been associated with non-domestic activity (e.g. in the Balfarg/Balbirnie complex).

The Grooved Ware finds from eastern and central Scotland include the assemblage from a pit at Littleour, Perth & Kinross (Sheridan 1998), which is associated with the latest date for Grooved Ware in Scotland (as cited in MacSween 2007, fig. 33.4). Prima facie the dates (from organic residues inside three pots, plus birch charcoal from their pit) suggest that Grooved Ware continued to be made into the second half of the third millennium, thus overlapping with the use of Beaker pottery. However, the three organic residues (from Pots 2, 3 and 6) have produced disparate dates ranging between 4110±55 BP (OxA-8992, 2880-2490 cal BC at 2σ, Pot 6) and 3845±75 BP (OxA-8993, 2550-2040 cal BC at 2σ, Pot 3), and the birch has produced an even later date of 3750±50 BP (AA-22906, 2340-1980 cal BC at 2σ). This spread of dates for material that was ostensibly deposited simultaneously is suspect, and re-dating, along with a critical review of the existing organic residue dates, is required.

Figure q: Ground plan of timber building at Stoneyfield, Raigmore, Highland. From Simpson 1996

Evidence for settlements is sparse and, once again, mostly takes the form of pits. A roughly rectangular post-built structure at Raigmore, Highland (Figure q; Simpson 1996; 1999), could conceivably have been a house but its nature, date, and association with the Grooved Ware pottery found at that site are all debatable.

Elsewhere, circular timber structures featuring an internal setting of four posts (Figure 39) are known from Greenbogs, Aberdeenshire (Noble et al. in press) with several others in Angus and Perth & Kinross known from aerial photographs - including a cluster of at least eight such structures at Chapelton, Angus (Noble et al. in press). The two Greenbogs structures have been radiocarbon-dated to between 2900 BC and 2500 BC; a single sherd of Grooved Ware may have been associated. While these resemble Grooved Ware houses from elsewhere in Britain (including Trelystan in Wales and Beckton in Dumfries & Galloway: ibid.), they also resemble Late Neolithic structures of ceremonial function (e.g. at Ballynahatty, Co. Antrim and Knowth, Co. Meath, Ireland) and others whose function could be either domestic or special purpose (e.g. at Machrie Moor and Durrington Walls). What is striking, however, is the similarity of form over large areas, together with the novelty of the design. Whether any of these structures evoked Late Neolithic structures in Orkney is open to debate, however. (These structures should not be confused with Iron Age structures featuring a square four-post setting, as seen for example at Grantown Road, Forres: Cook 2008.

Figure r: 'Four-poster' timber structures (A and B) at Greenbogs, Aberdeenshire. From Noble et al. in press

Evidence for funerary practices comes in the form of deposits of cremated bone. Recent dating of cremated bone that had probably been buried as foundation deposits in the sockets of the stone circle at Balbirnie, Fife (Gibson 2010) has shown that their deposition - and hence the construction of the stone circle - took place between the 31st and 29th century BC. At Orwell, Perth & Kinross, one of two deposits of cremated human bone placed in the stone-hole of a standing stone - one of a pair - immediately after its erection has produced a date of 4180±35 BP (SUERC-18309, 2890-2630 cal BC at 2σ: Ritchie 1974, 8; Sheridan 2008, 201). At Forteviot, Stirling, a cemetery containing at least nine discrete deposits of cremated human bone, plus extensive scatters, has been radiocarbon dated to the first three centuries of the third millennium BC (Noble & Brophy 2011a, table 1 - although note that not all the dates cited there have been calibrated to their 2σ value). The discrete deposits were found in shallow scoops, and the excavators suggested that they may have been in wooden box-like containers. One of the deposits may have been marked by a standing stone. One contained a fragment of a burnt bone pin, another, sherds of an undecorated pot. This cemetery is reminiscent of that excavated at Cairnpapple , West Lothian (Piggott 1948) although it would appear that the Cairnpapple cemetery is a little earlier, dating to within the last three centuries of the 4th millennium. (See below, At Raigmore, one pit contained the cremated remains of two adults, along with a petit tranchet derivative arrowhead (Simpson 1999). At Fordhouse Barrow, cremated human bone found in a putatively Beaker period pit (and quite possibly representing residual material) has been dated to 4340±35 BP (SUEC-2726, 3090-2890 cal BC at 2σ: Proudfoot 2004). Finally, at Culduthel, two Grooved Ware pots (one a tall narrow jar, the other a small pot) found in a pit surrounded by 'satellite' features were thought, by the excavator, to have been a grave, but the burnt bone found in the pit has not been confirmed as being human (context inf. Ross Murray).

Much more is known about ceremonial (and other special) structures, thanks largely to the excavations in the Balfarg/Balbirnie complex and, more recently, the excavations at Forteviot. These structures consist of:

  1. At least one stone circle (at Balbirnie, and probably also at Balfarg; and see below)
  2. A pair of standing stones at Orwell (although whether these had originally formed part of a larger monument is unclear)
  3. Several timber circles - at Balfarg, North Mains and Forteviot, in each case associated with a bank and ditch henge constructed several centuries later
  4. Other henges
  5. A large timber enclosure with an avenue-like entrance at Forteviot, along with a timber circle.

Also of note is the stones marked with rock art (i.e. cupmarks and cup-and-ring marks), which are relatively abundant in Tayside and Fife but rarer elsewhere in the region under review.

Each of these will be discussed briefly below.

The dating of the relatively small, low stone circle at Balbirnie (Fig. 56) has provided valuable confirmation that this type of monument was indeed being built in eastern Scotland around the beginning of the third millennium BC. Despite its structural differences with the Stones of Stenness in Orkney (which is known to have been erected, complete with its surrounding henge, during the 30th-29th century BC), it is conceivable that the inspiration for building the circle came from Orkney. This would accord with the fact that Grooved Ware was found in association with the circle at Balbirnie. (See Gibson 2010 for a discussion.)

Whether any other stone circles in eastern Scotland were constructed around this time is, however, a moot point, since the excavations of Richard Bradley at Clava cairns, recumbent stone circles and at the Broomend of Crichie henge (Bradley 2000; 2005; 2011) have demonstrated that all those monuments were erected after the conventional end-date for the Neolithic. (Indeed, in the case of Broomend of Crichie, the very concept that the stones had ever formed a circle has been convincingly dismantled: Bradley 2011, 75-80 and illus 2.9). Furthermore, his reconsideration of the Croft Moraig stone circles has concluded that the small oval setting (and, by extension, the numerous other oval stone settings in Tayside) was constructed during the second, or even perhaps the early first millennium BC (Bradley & Sheridan 2005. As for the date of the earlier, SE-orientated stone circle, this is to be explored during fieldwork in 2012.

The evidence from Balfarg henge (Figure s; Mercer 1981; Mercer et al. 1988; Gibson 2010) is more ephemeral than at Balbirnie, and it is not clear whether we are dealing with one stone circle, two, or some other form of stone setting. However, it has recently been argued (Gibson 2010, 72) that the stones were erected during the second quarter of the 3rd millennium, possibly to replace one or more timber circle, for which a terminus post quem of the 29th-25th century exists.

Figure s: Plan of timber circles at Balfarg henge. From Mercer 1981

The pair of stones at Orwell, with the deposit of cremated bone in the packing of one, has already been mentioned; it is unclear whether this had originally formed part of a larger monument such as a stone circle, or whether these represent a class of Late Neolithic monument - paired standing stones. Elsewhere in Scotland, it is clear that short stone rows are considerably later, dating to the Middle Bronze Age.

At Balfarg (Mercer 1981) and North Mains (Barclay 1983), the timber circles consist of multiple rings, not necessarily contemporary with each other (although at Balfarg, it has been suggested that the outer rings may have supported wattlework screens, restricting the view of the inner, free-standing post ring: Mercer 1981). At both sites, the henge bank and ditch was constructed outside the timber rings and at North Mains, the the date of 3665±45 BP (GrA-24007, 2200-1920 cal BC at 2σ), for cremated bone from burial A, under the henge bank, demonstrates that the henge was built several centuries after the timber circles (Sheridan 2003).

The timber circle at Forteviot (Figure t; Noble & Brophy 2011a) also predated the henge monument with which it is associated, but here the latter was constructed inside the timber circle. Both the circle and the henge surrounded the aforementioned cemetery of cremated remains deposits, and all these features were enclosed within the large 'palisade' enclosure discussed below. The radiocarbon dating of the circle (and the palisade) is based on oak charcoal but the excavators were careful not to use heartwood for dating, so with luck there should not be a considerable 'old wood' effect. The dates for the circle came out at c 2850-c 2450 cal BC (ibid., table 1) and are likely to post-date the cemetery; the dates for the henge (based on short-lived species charcoal) suggest that it was built between the 25th and 23rd centuries BC (ibid.). As for the 'palisade' enclosure, dates for its avenue-like entrance suggest that it was built between c 2900 and c 2450 BC (ibid.); Bayesian modelling would be necessary to produce a more precise estimate of the relative chronology and sequencing of all the structures here.

Figure t: Timber circle (along with other prehistoric features including a massive Late Neolithic 'palisaded' enclosure) at Forteviot, Perth & Kinross. From Noble & Brophy 2011a

A further example of a timber circle of possible Late Neolithic date, once more associated with a henge, is known from Moncreiffe, Perth & Kinross (Stewart 1985). It should be noted, however, that the 'Grooved Ware' from this site is much more likely to be Middle to Late Bronze Age 'flat-rimmed ware'.

Kirsty Millican's survey of the timber/pit circles of Scotland (Millican 2007) concluded that there are significant clusters along the southern edge of the Moray Firth and in Tayside and Fife. Whether all of these had been free-standing timber circles (as opposed to parts of round houses), and how many are of Late Neolithic (as opposed to Bronze Age or later) date, cannot however be determined without excavation.

The aerial photography record has also revealed the probable former existence of a number of henges in Tayside and Fife, in addition to the examples mentioned above (Dunwell & Ralston 2008, fig 9). These include a particularly large candidate at Westfield, Angus (Barclay 1999). Again, excavation is required to ground-test the evidence. It remains an open question, however, as to whether any henges in this part of Scotland were built prior to 2500 BC.

The large timber 'palisade' enclosure at Forteviot (Figure 18; Noble & Brophy 2011a, fig. 2; 2011b) is one of a set of similar, Late Neolithic enclosures, the others of which are known from Dunragit, Dumfries & Galloway; Meldon Bridge, Scottish Borders; Blackshouse Burn, South Lanarkshire (but in earthen form) and - from aerial photographs - Leadketty, Perth & Kinross (Noble & Brophy 2011b, fig. 2). The Leadketty enclosure is only 3km away from Forteviot in Strathearn. The possible functions of these enclosures have been discussed (e.g. by Noble & Brophy 2011b; Noble 2006; and Speak & Burgess 1999); all agree that they will have involved a large amount of effort in their construction and will probably have served as gathering places for large numbers of people. Whether they had been the focus for ceremonial activity, foreshadowing the later large henge monuments in southern England (e.g. Durrington Walls), is a moot point.

Finally, the number of 'rock art' sites in eastern Scotland has grown, largely to the painstaking survey work in Tayside and Fife by John Sherriff (Sherriff 1995). A considerable number is now known in Angus, especially around the source of the Lunan Water (Dunwell & Ralston 2008, fig 10).

Figure u: Edge-polished discoidal flint knife from near Huntly, Aberdeenshire. From Wickham-Jones 1987

The material culture record for Late Neolithic east and central Scotland - in addition to the Grooved Ware pottery mentioned above - is dominated by carved stone balls (Figure 75) and maceheads (Figure 74). Over 80% of all the 500+ known carved stone balls have been found in north-east Scotland, and even though a few may have acquired an 'Aberdeenshire' provenance due to antiquarian collecting practices during the 19th century, nevertheless this concentration seems genuine, and it is likely that they had been made using local, carefully selected erratic cobbles. Both these and the maceheads form part of the overall assemblage of Late Neolithic special-purpose and symbolically-significant items, which would have served as symbols of power and prestige, as well as perhaps having other functions.

As for other lithic items, while it appears that pitchstone was probably no longer being imported to this particular part of Scotland at the time (Ballin 2011, 3), the link with Yorkshire remained strong, with items such as the edge-polished discoidal knife found near Huntly, Aberdeenshire (Figure u; Wickham-Jones 1987 and cf. Clark 1932) being probable imports. Grooved Ware pottery also shows that design ideas were being shared with Yorkshire (and indeed elsewhere in England) as shown, for example, in the near-identity (Figure v) of a pot from Midmill, Aberdeenshire and one from North Carnaby Temple Site 2, Yorkshire (Manby 1974, fig. 18).

Figure v: Very similar Grooved Ware pots found at Midmill, Aberdeenshire (left) and North Carnaby Temple 2, Yorkshire (right). Midmill reproduced courtesy of Murray Archaeological Services; North Carnaby Temple image from Manby 1974 Research questions regarding Neolithic east and central Scotland

The principal questions concern the need to improve current knowledge of the Neolithic in this part of Scotland, and to integrate all the strands of evidence reviewed above (and in Theme 4) into an overall narrative, as follows:

  1. What is the overall pattern of settlement, land use and subsistence activity and how and why does it change? This will need the compilation of period-specific, GIS-based distribution maps of all the evidence already available, and this will involve trawling through the grey literature.
  2. What was the nature of Middle and Late Neolithic habitation structures? In other words, are we dealing with a change in building techniques, away from the Early Neolithic rectangular, post/post and plank-built construction (as the Late Neolithic 'four poster' structures suggest) - and were these structures used in a different way from Early Neolithic houses?
  3. What can be said about the nature of, and changes in, society from the evidence currently available? Clearly this part of Scotland became involved in the larger social changes from c 3000 BC onwards, but did the nature of society change? Why was this region the epicentre for the production and use of carved stone balls? How exactly were the large 'palisade' enclosures used? And how does the practice of creating and using rock art articulate with other expressions of belief and ritual practice?

There are also more specific questions, including:

  1. How many of the timber/pit circles and henges or henge-like sites known from the aerial record actually belong to the Late Neolithic?
  2. Relatively little is known about the use of animals in this part of Scotland, since bone assemblages do not survive well in the free-draining soils. There is a need to find many more, and larger, assemblages in order to improve our understanding of species exploitation and management techniques.
  3. What prompted the initiation of flint mining on the Buchan Ridge Gravels? Could it have been a reaction to the existence of a flourishing specialist flintworking tradition in North Yorkshire?
  4. The Littleour Grooved Ware assemblage needs to be re-dated, and the existing dates critically re-evaluated; and there is a general need to round up all the existing dating information about all Neolithic pottery in this region, to refine our developmental sequence.
  5. The dating of Neolithic non-megalithic round mounds needs to be improved, and unless the Early Neolithic cremated bone from Pitnacree turns up in the Duckworth Laboratory in Cambridge University, the only way to achieve this will be through excavation. The round mound at one end of the Cleaven Dyke would be a particularly interesting monument to explore, especially to examine its relationship (if any) with the Cleaven Dyke.
  6. More examples of the 'four post' timber circle structures need to be excavated, to investigate their nature, use and relationship with other similarly-shaped structures elsewhere.
  7. How many of the stone circles in this region date to the Neolithic? In particular, might the earliest stone circle at Croft Moraig be contemporary with Balbirnie?
  8. Are paired standing stones a Late Neolithic phenomenon - and if so, is it restricted to Tayside and Fife?
  9. The nature and date of the activities at Stoneyfield, Raigmore, need to be clarified through a programme of radiocarbon dating.
  10. The use and significance of cursus monuments and bank 'barrows' needs to be clarified.

5.1.1 Early Neolithic, to c 3600 BC

Essentially, the story starts with the appearance of two north French ceramic traditions:

  1. a Breton tradition, featuring Late Castellic and related pottery, which appears in the west of Scotland (at Achnacreebeag) as part of the 'Breton, Atlantic' strand of Neolithisation and
  2. a North-French tradition, constituting one of a variety of regional styles of 'Chasséo-Michelsberg' pottery and which is known in Britain and Ireland as 'Carinated Bowl' pottery . (The justification for using this term to describe a tradition that encompasses non-carinated forms as well as carinated forms is explained in 2007; and note that there is still variability in other people's use of terminology, with some still using the obsolete 'Grimston' or 'Grimston-Lyles Hill' terms, or the over-vague term 'bowl (or Bowl) pottery'. This is to be discouraged as it leads to confusion.)

The origins of the Breton tradition lie in the Morbihan region of south-east Brittany, while the Carinated Bowl tradition is most likely to have originated in the Nord-Pas de Calais region of northern France, where excavations at Étaples have produced a close comparandum.

The distribution of these two traditions, in their initial form, does not overlap; the former is limited to the west of Scotland, while Carinated Bowl pottery use extends over much of southern and eastern Scotland, as far north as Caithness. As will be seen below, however, in its later variants there was 'cross-fertilisation' between the two traditions as potters shared design ideas. It should also be noted that there were parts of Scotland - the Outer Hebrides (and some of the Inner Hebrides), north-west Mainland, parts of western mainland Scotland and the Northern Isles - where pottery did not begin to be used until the late 38th or 37th century BC.

Breton-style pottery

The Breton tradition is characterised by the use of thin-walled, fine-textured bipartite bowls (of Late Castellic style) with distinctive decoration: the example from Achnacreebeag has a 'rainbow' motif above the carination and a fringe of short vertical lines below. Also present in the simple passage tomb at Achnacreebeag were sherds of two other pots of Breton style, more simply decorated with stab designs. The Breton parentage of this pottery is very clear, with comparanda, for example, from a simple passage tomb at Vierville, Normandy - a pot which represents the northward movement, either of the vessel itself or of the ceramic style, from Normandy (Cassen 2011 ) - and from the funerary complex at Locmariaquer in the Morbihan area of SE Brittany (ibid.). The development and dating of Late Castellic pottery is described in Cassen et al. 2011 , from which it is clear that the assemblage at Achnacreebeag dates to between c 4300 BC and c 4000 BC. This is therefore the earliest pottery in Britain and Ireland, standing at the start of a tradition of using decorated bipartite bowls in the west of Scotland and in Ireland: later examples of this tradition, showing clear stylistic 'parentage' from its Late Castellic origins, are to be found in the Clyde cairns of SW Scotland (where Jack Scott termed them 'Beacharra Bowls') and in the court tombs in the north of Ireland (where Humphrey Case called them 'Ballyalton bowls': 1961). The longer-term development of this particular ceramic style has been traced by Sheridan (1995; 2003), where its persistence in Ireland to around the 36th century BC can be traced .

Petrological thin-sectioning of sherds from the Achnacreebeag pots (by Gwenaëlle Hamon, for Alison Sheridan), and comparison with Breton Late Castellic pottery, has unfortunately proved inconclusive in determining whether the former had been actual imports to Scotland; the fabric is so fine that there are too few lithic inclusions to allow a consideration of origins on that basis. However, further work on examining local clays may help in this enquiry. It seems likely, however, that it was the know-how to make this pottery, rather than the pots themselves, which moved from Brittany to Scotland. This will have been a totally new technology in late 5th millennium Scotland, and the skill with which the Achnacreebeag assemblage had been made means that it can only have been made by a skilled and experienced potter/s.

It should be noted that the objections that have been raised to the idea of a Breton ceramic tradition having been introduced to Scotland (Whittle et al. 2011, 808ff) are based on a misunderstanding of the ceramic sequence in Scotland, with the later versions of this pottery type (e.g. at Beacharra) being assumed to be the 'parent' of the Achnacreebeag pots. This is an error (as explained in Sheridan 2012), and in any case no plausible alternative explanation for the appearance of an entire ceramic tradition has been offered.

The Carinated Bowl tradition

This tradition features the use of carinated and uncarinated vessel forms, all undecorated (save for the occasional use of decorative finger fluting ). Sheridan has made a distinction between the earliest manifestation of this ceramic tradition, which she has called 'traditional Carinated Bowl [henceforth CB]', and subsequent developments, called 'modified (or developed) CB' (2007; note the importance of using capital letters to distinguish between the tradition and the vessel form). 'Traditional CB' pottery is markedly consistent (in fabric, form and finish) over a wide area in Britain and Ireland, whereas 'modified CB' shows regional variation as the process of 'style drift' led to changes, in different ways, at different rates, in different areas.

Large Carinated Bowl pot from Knocknab, Glenluce, Dumfries & Galloway. Photo: Alison Sheridan

'Traditional Carinated Bowl' pottery features carinated bowls in a variety of sizes and shapes , mostly (but not exclusively) fine-textured, and mostly thin-walled. Some are extremely thin - as thin as 4 mm in some cases - and the whole tradition is the product of skilled potters with over a millennium of experience in making pottery. As with the Breton tradition, its appearance in Scotland, probably during the 39th century BC, represents a wholly alien technology, and the skill with which this pottery was made shows that it was made by people who were used to making pottery, and who followed conventions in the 'recipe' and techniques of pot construction as well as in the shapes to be made.

The carinated forms range from broad, shallow bowls to deep-bellied bowls; one of the latter, from Auchategan, Argyll and Bute, is of classic Michelsberg 'tulip beaker' form while others would seem 'quite at home' in Chasseo-Michelsberg assemblages from northern France. Carinations are generally gentle, and in some cases the pots are S-profiled. The uncarinated forms comprise generally roughly hemispherical bowls and cups, and also large necked jars, with a deep-bellied, S-shaped profile. With all the pots, surfaces have been carefully smoothed, and in some cases the exterior (and sometimes interior) have been polished to a low to medium sheen, or even burnished to a high sheen. Lithic inclusions are sparse, often less than 3% in density, and often feature the use of crushed granitic stone, with tiny mica platelets giving the surface a slight glitter. Widespread technical details of manufacture include the marked thinning of the neck just above the carination, which is particularly prevalent on the wide shallow bowls; while this weakened the vessel - with many pots having broken at this point - it nevertheless helped in achieving the desired vessel shape. Once more, the widespread occurrence of this feature points strongly to CB pottery having been introduced by potters who shared the same ceramic tradition.

While some have accepted Andrew Herne's suggestion that carinated bowls had a special, ceremonial function, the evidence does not substantiate such a view. There is abundant evidence for the use of some carinated (and a few uncarinated) bowls for cooking, but they would also have been used for serving, and larger vessels for storage. This pottery has been found in the full range of contexts, from the domestic to funerary and other monuments.

The subsequent development of this tradition is, as indicated above, a story of regional diversification, with a process of style drift having operated in different ways, at different times, in different areas. In north-east Scotland, as Henshall had observed as long ago as 1983, a regional style emerged early on, featuring the increased use of fingertip fluting and of ripple-burnishing; the addition of lugs; and the addition of new forms, including baggy lugged bowls. This 'North East Carinated Bowl' (NECB) pottery finds echoes in the developed Carinated Bowl pottery of northern Ireland (in what Case called 'Lyles Hill ware: 1961), and this is no coincidence, as there are documented links between NE Scotland and NE Ireland, via the Great Glen (e.g. in the use of porcellanite axeheads from Co. Antrim). The early date at which this 'style drift' occurred is indicated by the assemblage from the so-called 'hall' at Balbridie, Aberdeenshire: built within a generation or so of the near-identical 'hall' at Warren Field, Crathes, just across the River Dee, its assemblage is of NECB pottery, whereas that at Crathes is 'traditional Carinated Bowl'. The Balbridie assemblage includes sherds from two sharply carinated bowls with decoration on their collars : these constitute the beginnings of what was to become the Unstan Bowl, and it is assumed that this represented innovation in pot design.

Other regional versions of 'developed Carinated Bowl' pottery include a general coarsening and thickening of the vessels, the occasional addition of lugs, and shape deviation from the earliest carinated forms.

In the west and south-west of Scotland, a fusion of the Breton and Carinated Bowl traditions can be seen, with decorated bipartite bowls occurring along with developed CB forms to constitute Scott's 'Beacharra' tradition, with its regional variability. Into this mix was added the use of deep baggy lugged bowls; as argued elsewhere (e.g. Sheridan 2004), this may well represent the adoption of a south-west English style of pottery (formerly known as 'Hembury Ware', thanks to north-south interactions within the Irish Sea area from the c 37th century. Some sharing of styles between NECB pottery and that found in W and SW Scotland is clear from the distinctive pots formerly known as 'Achnacree Bowls' (and as part of 'Rothesay Ware' ): these have long, vertical necks, heavy, hooked rims and shallow bellies, and are decorated with vertical lines of either ripple burnish or incision. Examples have been found at Culduthel , near Inverness, in a domestic context; in the passage tomb at Achnacree ; in the Clyde tomb at Nether Largie in the Kilmartin Glen ; and in Clyde tombs in Bute. The example at Culduthel comes from a context radiocarbon dated to c 3600-3500 BC and this provides a reasonable estimate of the date of this pottery.

A recent attempt to use Bayesian modelling to characterise the dating of modified Carinated Bowl pottery in Scotland has been made by Whittle et al. (2011), but this has failed to take into account that defining an end to this process is difficult, since ceramic traditions tend not to end abruptly; instead, we can trace a gradual process of style drift as conventions change and as new design ideas are adopted, through local innovation and through links with other areas. For this reason the use of Bayesian modelling is flawed. (See Sheridan 2012).

The spread of pottery using to the Outer Hebrides and the Northern Isles

It was this fusion of Carinated Bowl and Breton-origin pottery that spread to the Outer Hebrides, parts of western Scotland and north-west Scotland, probably during the late 38th or 37th century BC. Jack Scott (1966) has termed this pottery 'Hebridean Beacharra', and examples can be seen from sites such as Northton (Simpson et al. 2006), Calanais (Ashmore forthcoming), Eilean Domhnuill (Armit 2003) and various chamber tombs (Henshall 1972). The repertoire includes both undecorated and highly decorated vessel forms, in a variety of shapes. The former include simple uncarinated bowls, bowls with flanged rims and deep lugged jars (with their 'Hembury/South-Western pottery' affinities) and the latter include large ridged jars and Unstan Bowls - a type of vessel shared with Orkney and north-east/east Scotland, and almost certainly adopted in the Hebrides thanks to contacts with Orkney and/or the north-east mainland. Decoration is almost exclusively by incision. The dating of this 'Hebridean Neolithic pottery' tradition leaves much to be desired, with the existing dates suggesting an implausibly long currency from around the 37th century (or even possibly late 38th century) until the early third millennium. Improvement of the dating is a key research priority.

As far as the earliest Neolithic pottery in Orkney is concerned, this too contains an undecorated and a decorated component, with Unstan Bowls possibly being present from the beginning. Close ceramic links with the north-east mainland are in evidence, and it may be that the farming communities who pioneered farming in Orkney came from the north-east Mainland. The widespread use of the term 'Unstan Ware' offers a misleading impression of consistency among the pottery associated with Unstan Ware, and should be avoided. Key assemblages for understanding the early development of pottery include the Knap of Howar (Ritchie 1983) and Pool (Hunter 2007).

The earliest pottery in Shetland consists of a handful of very small, undecorated, slightly coarse sherds from West Voe. While these are too small to be diagnostic, they are most likely to resemble the plainware component of the ceramic tradition of the Hebrides and western mainland.