The story of the Neolithic period in Britain as we so far understand it has been compiled from myriad individual archaeological encounters with the traces of human activity from the centuries concerned in different places within the landscape. These traces include the remains of partly earth-fast timber structures which often consist of recognizable features representing where the timbers had been pulled out of the ground, or had rotted in situ, or had been burned; areas of burning of ground-surfaces where hearth-fires had been laid; spreads of decayed materials that were formerly rubbish dumps or ‘middens’; large holes (usually referred to as ‘pits’) dug and backfilled with various deposits including whole or broken artefacts thrown or placed within them; and ditches that had been infilled or had silted up, and sometimes re-dug and redefined. The different episodes of construction and deposition that led to the formation of these traces are differentiated by those investigating them through the identification of thousands of isolable ‘events’. Some of these events were almost momentary (the digging of a pit, the removal of a post), while others (such as the gradual silting of a ditch) took place over an extended period. Archaeologists describe the isolable actions, events, and deposits resulting from such occupation of the land as ‘contexts’. Some materials retrieved from some of these contexts have been carefully selected by the archaeologists during their investigations to be datable using a variety of scientific dating techniques, and they provide individual site chronologies linked closely to the stratigraphic sequences involved. Repeated observed associations of different kinds of artefact with reliably dated contexts and site sequences allow comparative chronologies to be painstakingly constructed, and it is from this process that the possibility of a chronologically sound historical narrative for the Neolithic is gradually being built up.
In these chapters entitled ‘narratives’, however, we are not only talking about the sketching out of a historical sequence, extremely important though it is. We are speaking also, if to a limited extent, about the teasing out of multiplicities of story from the material evidence. So, the presence and course of individual lives is immanent in the individually made, broken, and placed items, and the residues of immediate action, that are encountered. Equally, the places of habitation each have their own narrative that encapsulates aspects of the history of the communities that created them and extended their use-life through time. Moreover, the stories of interaction between communities are embodied in the locales, the substances, and the made items through which social relations, interactions, and exchanges were enacted.
The fourth millennium bce in Britain: a broad-brush perspective
One way of perceiving the history of the fourth millennium bce in Britain is to see it in terms of a big picture comprising the grand narrative of the arrival and spread of a Neolithic ‘package’, along with Continental culture-bearers. As we have seen, this is much as David Miles has recently proposed in his book The Tale of the Axe. While this can be understood basically as a social evolutionary approach, it has become nuanced by a new attention to the transformation in world views as well as subsistence practices that occurred across the span of time involved, and it has been finessed in respect to new fine-scale chronology. It is more generally understood, and is acknowledged by Miles and others, that there is manifestly much variation among the sites and artefacts encountered from the first centuries of the Neolithic period in Britain. It is now widely acknowledged that this must have implications for our awareness of regional trajectories of change and the processes of cultural differentiation that were likely to have occurred at different points in the millennium concerned. Curiously, such writers appear to have brought us full circle, since it is clear that this much was recognized also by Stuart Piggott back in 1954. So how far have we really moved on, interpretively, in the intervening sixty years?
Despite the evident variability, there are clearly a number of themes in common across the centuries concerned from region to region, and also some broad horizons of change. These horizons are registered in the adoption of practices that recurred across sometimes (we now understand) relatively short spans of time, and in the subsequent ‘referencing’ of earlier activity by later activity in the same locations. A first horizon, for example, saw the appearance of some of the earliest manifestations of Neolithic practice, which occurred within a landscape that had perhaps been more extensively occupied by a late Mesolithic population than is sometimes acknowledged, and often employing ancestral locations. These phenomena included sequences of acts that would culminate in the construction of some of the earliest long barrows, fine bowl pottery, pit-digging and deposition, polished flint and stone axes, individual inhumation burials, and other distinctive workings and structures (including some of the Sussex flint mines), large and complex pits, timber ‘halls’, and some houses, in the period between around 4050 and 3750 bce. The decades around 3800 bce saw these same entities rapidly appearing for the first time in the north of England, southern Scotland, Wales, and the south-west peninsula, even as some of the practices concerned began to wane and to be supplanted by new innovations in the south-east.
A second horizon is marked, broadly from 3750 to 3600 bce (with again the earliest dates from the south of England), by the appearance in pits and other locations of more items than previously that had travelled over some distances (including, only now, pots from the south-west peninsula), and novel kinds of site including, especially, the distinctive ‘causewayed enclosures’ discussed at some length later in this chapter. Meanwhile, other structures such as the early halls ceased to be built, and the production and use of fine carinated pots gradually declined. The latter were supplemented and before long apparently supplanted by the use of distinctive shouldered vessels and deep bag-shaped bowls in a variety of coarser fabrics. From around 3700 bce, long mounds became more numerous and more varied in form, with whole traditions of construction mushrooming across the British mainland. As causewayed enclosures became more numerous, and the repertoire of deposition more varied, the forms of pottery in use included an increasing variety of surface decoration. In the earlier part of this ‘phase’ of activity, kinds of site not previously seen, such as post-defined linear monuments, made their appearance in Scotland.
A third horizon is represented during a later stage in the use of causewayed enclosures from 3500 to 3300 bce by a pronounced shift in the pattern of ‘exchange’ linkages, most dramatically evident at sites such as Hambledon Hill in Dorset. In the same period, the enclosures themselves were often reconfigured and reused, and long mounds were often monumentalized before being abandoned. Coarse pottery became yet more elaborate and highly decorated, and signs of warfare occur at several sites, with cases of trauma showing up in skeletal remains. Parallel-sided embanked cursus monuments became common in many parts of southern Britain, in several instances built across the abandoned earthworks of causewayed enclosures. Further changes ushered in a period that extended down to 3100 bce and beyond, during which in some places individual burial was monumentalized while elsewhere timber ‘palisaded’ enclosures and cremation cemeteries appeared for the first time.
This rapid overview has necessarily suppressed much of the complexity and variability of fourth-millennium history. The rest of this chapter is therefore devoted to exploring something of that richness as revealed through different dimensions of the evidence and as expressed here through a series of overlapping narratives vested in contrasting traditions of practice. Inevitably, we focus most closely on those locations, kinds of site, and artefacts, the remains of which have survived most extensively and are best-preserved and more frequently encountered; although exceptional survivals and some individual remains and artefacts provide insights disproportionate to their number.
In the beginning, there was fine pottery...
As we have already argued, the story of the first efflorescence of the Neolithic in Britain is one characterized by relatively abrupt change in cultural practices against a background of continuity, both insular and Continental. The transformations that took place within Britain across the span of time running from the last centuries of the fifth millennium through to the first quarter of the fourth millennium bce can be identified as much with the rapid development of a radically different way of defining community as with a switching of mode of subsistence. In constructing a narrative for this period, we will inevitably consider in more detail some of the strands of cultural practice that emerged, and identify the various trajectories of change that can be traced through the first centuries of the fourth millennium. Such a narrative takes us onwards to an era in which the interactions and linkages between communities became much more clearly manifest, and the tempo of change appears to have quickened.
Some of the earliest dated traces of the people whom we could identify as ‘Neolithic’ are found in contexts that would at any point and in any place in the following 2,000 years be regarded as unusual. So, for example, by the side of the River Thames at Yabsley Street, Blackwall, downstream from modern London, a timber-lined elongated pit had been dug most likely sometime in the period 4028–3990 cal bce (as mentioned in Chapter 2). The pit was evidently a coffin-like grave in which the body of a woman had been placed. Close to her head had been placed a large sherd from a fine, angular-shouldered, thin-fabric, highly burnished carinated bowl, while pieces from further vessels were found elsewhere in the grave. Near St Albans the cremated remains of another adult had been placed, sometime around 3900 bce, in a hollowed-out canoe-like tree trunk at least five metres long within a grave full of traces of burning. On Cranborne Chase near Blandford in Dorset and near Maidenhead in Berkshire, deep shafts have been found, dated to shortly after 4000 bce, into which pieces from similar pots, worked flints, and in one case fragments of stone axes had been tipped.
One of the most common kinds of location from which such early pottery has come, however, is the ashy organic spreads of debris that we have already referred to as middens. The character of such accumulations, in which apparent food waste (mostly animal bone), broken pottery, and worked (and sometimes burnt) flint are regularly found, has been much debated. For instance, were these objects incidental to domestic activity, or were they being deliberately curated (that is, purposefully stored within maintained piles from which they could later be taken and reused)? Certainly, there is growing evidence that material of various kinds had been retrieved from these piles after a period, and then intentionally re-entered into circulation. It also appears that sequences of deposition and the creation of structures within and around which such material accumulated were a recurring feature: in some cases these were light timber structures, and in others stone-slab defined boxes.
Elsewhere, parts of several pottery vessels, including five carinated bowls, were found in two successive fills of a substantial Early Neolithic pit at Eweford West, near Dunbar, to the east of Edinburgh. Close by, a mound of turves and earth represented an enlargement of a primary mound and pit, and over this turf mound had been spread individual pieces of similar pottery. Timber-lined chambers that had been burnt, and that were associated with human remains, were then inserted on either side of this mound. In this way, spreads of material usually thought of as accumulated ‘occupation debris’, often apparently from deliberately reworked middens, recur at a series of different locations. While such occurrences have in most cases been dated, the principal place accorded them within conventional narratives of ‘the development or spread of the Neolithic’ has been as indicators of a stage reached in the process of ‘becoming Neolithic’. We would consider them instead to represent instances of lived practice in the past. Each example constitutes part of the unfolding history of these practices, which cumulatively brought about the transformation of the social worlds that people inhabited.
However, our awareness of time and its passing during the Neolithic has been transformed by the development of finer chronologies. As we have seen, these have often been based on the Bayesian statistical approach that brings together absolute dates and finely scrutinized contextual information: identifying whether the sampled material was encountered in its original place of deposition, for instance. While the dates concerned can still be used to ‘map’ the adoption of Neolithic ways of life, they can also facilitate a focus upon the cultural significance of different social practices and their material outcomes. One aspect that we have already touched upon is the role of ‘the house’ in the establishment of distinctly Neolithic communities, and a sense of collective identity amongst the groups concerned. We would argue that this involved rather more than an awareness of ‘being Neolithic’, and indeed there is no guarantee that such a notion would have been comprehensible to these people. Instead, we suggest that a new way of life was facilitated by the formation of what one of us has previously referred to as ‘house societies’, and for which we will here coin the neologism of ‘invested lineages’.
A narrative of Neolithization, and the creation of houses or halls
As more dates from a variety of contexts have accumulated, and more particularly since it has become possible to model these convincingly using both sophisticated statistical techniques and stratigraphic arguments, it has been possible to trace the introduction of Neolithic innovations and practices as a process that began in the Thames Estuary region from as early as 4100 bce before expanding over the subsequent three centuries towards the outer limits of the mainland of Britain. This is a model, and by definition a simplification of reality. The danger in expressing such an argument in graphic form, as a map of the ‘spread of the Neolithic’, is that it appears to be based upon data that is entirely solid. But in reality, it is based upon indicative data, which means that it glosses considerable disparities in the amount of reliable information available for different regions. This point is explicitly acknowledged in the Gathering Time study, but there is a danger that the results might be employed without acknowledging its limitations. Most notably, there are crucial lacunae that to some degree compromise its general utility beyond the comparatively data-rich areas: in several parts of the British mainland, there is a notable paucity of reliable evidence. In the north of England, meticulous modelling has been undertaken by Seren Griffiths with important results, but the overall number of dates on which it is based remains modest. Still more limited is the data for Wales and the Marches, which relates to only a handful of sites, and as we shall note later in respect of places such as Herefordshire, such areas will undoubtedly yield dating ‘surprises’ in the years to come.
We could arguably produce some form of alternative rendering that attempted to express these uncertainties in a visual form, but the result would probably still mask the lived complexities and reinforce the impression of a straightforward spatial and temporal progression. The messy reality of lived experience alternatively informs us that such simplicity is the exception and not the rule, that the spread or adoption of change is only very rarely a linear process, and that ‘Neolithization’ will not inevitably occur in a ‘wave-like’ manner, as proposed by Ammerman and Cavalli-Sforza back in the 1970s. A variety of factors will condition the dispersal of things and practices (and their transformation in the process), ranging from topography, vegetation, and established routeways, to collective knowledgeability, sociality, and the interconnectedness of communities across sometimes considerable distances, separated by other groups that do not necessarily share the same networks of contact. As this process was lived out, it would produce a variegated pattern in the emergence of Neolithic societies, and we should not be surprised to find that the adoption, introduction, or adaptation of new ways of living leapfrogged its way across the landscape. This is important, because advocates of the wave model of change will tend to attribute an early date for ‘Neolithic’ activity from a site, say, in south Wales, to scientific, sampling, or investigatory ‘error’, because it does not fit the preferred model of orderly spatial progress. This is a much simpler, if less rigorous, reaction than to take the evidence in the first instance at face value, and accept that it indicates that the process was both dynamic and stochastic. The precise pattern of appearance, adoption, or change is likely to have been affected to some extent by links of various kinds that preceded the commencement of the adoption of farming and/or the first appearance of Neolithic artefacts. These connections might include consanguinity and kinship, or networks facilitating the circulation of objects or information.
It seems possible that in the fullness of time we will be able to offer a more sophisticated account of the dispersal of the Neolithic in Britain, but for the moment there are several other narrative strands that we might follow through this earliest Neolithic period. We will explore just two of these that appear to us to be significant, but first we should pause and note that the exploration and interpretation of the earliest Neolithic of Britain also have the capacity in future to change the nature of the question itself. It is open to question how far ‘arrival and spread’ will remain the most important dimension of narratives of Neolithization, and it might be concluded that this is a rather old-fashioned way of thinking. As such it may be replaced with one that focuses not upon a totalizing process as much as upon the ways that innovations were introduced locally, promoting specific ecological, economic, and social trajectories in the centuries that followed.
We would suggest that it is through an exploration of these various alternative narrative subjects that we can begin to see the outlines of just such a more historically literate kind of approach to the earliest period of Neolithic history. The first of the two subjects we have chosen to highlight here is the creation of ‘houses’ in both the physical and the social sense. The focus is not so much upon the relative dating of the initial construction and subsequent use of the different examples excavated in recent years, which we have already addressed in Chapter 2. Rather, it is what their building, and in particular their narrow temporal horizon, can tell us about their significance in the broader historical sense. The second subject concerns burial and the creation and use of bone repositories, and the historical moment encapsulated in the creation of the first examples.
We have already introduced the large timber halls that have been identified in southern Britain and in lowland Scotland and north Wales, but we have not yet expanded upon their character and significance, beyond emphasizing that they were both an early and a distinctive phenomenon. Not all of the early halls had the aisled layout that we have observed at White Horse Stone, but all were rectangular or sub-rectangular in plan (Fig. 3.2).
Most were post-built, but some may have had solid wooden side-walls comprising upright timbers, and at Dorstone Hill in Herefordshire the walls were apparently infilled with dung- and clay-based daub. These halls at Dorstone were deliberately burnt down, and this was also the case at Claish Farm, at Lockerbie Academy, at Doon Hill, and at Warren Field and Balbridie (on opposite sides of the Dee valley west of Aberdeen). This burning was not ubiquitous, however: similar buildings that were not burnt have been excavated at Yarnton in the Thames Valley, at Llandygai in north Wales, and at Lismore Fields near Buxton in Derbyshire.
One aspect of the building of these halls that may be significant is that there had sometimes been more than one such structure (or at Horton near Reading, smaller rectangular, post-defined buildings referred to as ‘houses’) at a location. It is rarely the case that the footprints of different buildings are superimposed upon each other, so it is difficult to assess whether these structures had been constructed sequentially, or had existed contemporaneously. Thus chronological succession is possible, but so too is social segmentation, with the buildings having been ‘owned’ by different kin- or clan-groups. Multiple structures have been recorded at Llandygai near Bangor, at Llanfaethlu on Anglesey, at Lismore Fields in Derbyshire, and now at Dorstone Hill in Herefordshire; and possibly also at White Horse Stone in Kent (Fig. 3.3).
As previously explained, the period during which such structures were built and used was relatively brief. In any one place where they occur, the span of their construction and use can probably be fitted within a period of anywhere from less than a decade to 150 years. So the refined chronology that we are now developing indicates that there appears to have been a time in each region when no such buildings were known, followed by a brief interval in which they were built and used, before the practice was again abandoned. In the comparable situation in Ireland, where the structures are much more numerous than in Britain but are predominantly smaller houses rather than large halls, all may have been constructed within a period of fifty-five to ninety-five years, a generation or two after the first Neolithic activity there.
This brevity of use may mean that the halls and houses concerned had a specific role (or roles) at a particular point in time, and that subsequently their construction ceased to be important to the communities concerned, or their purpose was met by other means. One way of explaining this is to argue that the craft-skills of the initial immigrant communities in building such structures died out, so that the successor communities were no longer capable of raising such sophisticated buildings. Alternatively, it could be maintained that living together in robustly constructed shelters might have been an important source of collective security for newly arrived colonists. However, the people in Britain responsible for the earliest structures were soon building other structures in wood, earth, and stone that were at least as sophisticated as these ‘foundational’ halls and houses. Moreover, as we saw in Chapter 2, with the possible exception of the site at Mairy in the Ardennes region of Belgium, closely similar structures are uncommon amongst communities on the near-Continent at the relevant period. It follows that the reasons for the rise and fall of hall-building should be found in the immediate conditions of the first Neolithic communities of Britain, whether we consider these to have been migrants, indigenous people, or some combination of the two.
So rather than being a circumstance forced upon these early groups, the innovation and the cessation of building such large and sophisticated structures were deliberate acts, intended to have specific outcomes. After a period of some decades it was apparently no longer necessary or appropriate to raise such structures. It is therefore our strong inference that building the house was at once an undertaking that drew people together in an act of labour, and one that established their solidarity and collective identity in a tangible and enduring form, building the household. The physical act of creating the structure bound the builders together into a new kind of community, which anthropologists refer to as a ‘house society’. This is a kind of social unit in which the right of residence in a dwelling structure, and shared ownership over resources, artefacts, and intangible goods such as names and rituals, define who is and who is not a member. This is not to say that these Neolithic halls and houses had only symbolic importance. They undoubtedly had practical uses as places for living (whether permanently or seasonally), for meeting, and for storage. But their sometimes impressive size and robust construction and their concentration in a particular horizon of time indicates that they encapsulated the idea of the inception and assembly of a new kind of community, whatever the origins of its individual members. In particular, we would emphasize that during this period the communities concerned defined themselves as entities capable of collectively owning and disposing of wealth. This was an arrangement that was essential both to facilitate the cultivation of plant foods and the keeping of stock, and to enable the accumulation and inheritance of resources between the generations. These groups thereby became what we have termed ‘invested lineages’, insofar as investment in capital projects (at this stage, halls and houses), and the collective organization of subsistence production, became fundamental features of their existence. Although the process itself occurred across a brief span of time, once these social and material investments had been made, and the necessary relations of gifting and intergenerational transfer had been entered into, it was impossible to go back to looser social and material arrangements. This question of the loss of options is one that it is important to understand in reference to a new ‘Neolithic’ way of comprehending the passage of history, and it is a point that we shall return to in particular in Chapter 6.
A narrative of human burial, and the first creation of bone repositories
Earthen long barrows and ‘megalithic tombs’ (the latter better described as barrows or cairns that featured the ‘architectural’ deployment of large stones) are arguably the most distinctive monuments of the Earlier Neolithic, not least because of the use in many of them of whole or split tree trunks or unworked boulders as major structural components (Fig. 3.4).
As we have related in Chapter 1, it was once conjectured that these monuments were created from the beginning of the period as territorial markers that were dominant and durable presences in the landscape. However, we now know that the earliest structures built in part at least to contain ancestral human remains were not always as prominent as they later became. To begin with, these barrows were few in number relative to their later ubiquity in some regions at least, and while the earliest structures possessed elements that prefigured later elaborations, some of these were relatively small in scale, even though they were occasionally strikingly complex.
Only a small minority of these sites have produced series of dates that place their beginnings clearly in the period before the thirty-eighth century bce. The stone-lined chamber set within the earthen barrow at Coldrum near the Medway river in Kent has possibly the best claim to be the earliest such structure so far known, perhaps dating to sometime in the fortieth or thirty-ninth century bce, with the circular stone tomb at Broadsands near Paignton in Devon being built a century or more afterwards. Potentially early dates have also been returned on bones from stone cairns in the Cotswolds at Burn Ground and Sale’s Lot, although the margin of error is sufficiently wide that either might have been constructed around 3800 bce or later, along with cairns at Notgrove and West Tump. Nearby, the earthen so-called ‘Banana barrow’ sealed beneath the bank of the later causewayed enclosure at Crickley Hill may also be early, although its precise character is unclear. An earthen mound at West Cotton in the Nene valley in Northamptonshire appears to have been some form of burial site or cenotaph that has also produced early dates.
All these sites, apart from the latter two, were the subject of early excavations; and for this reason, where the radiocarbon dates have been obtained from human remains the contexts from which they derive are not as clearly related to the construction of the monument as we might prefer. However, all the bones appear to have been associated with primary structures at each site, even if, as in the case of Burn Ground, it is possible that the human remains had been curated in some other location before ‘final’ deposition in the contexts in which they were found. The site at Yabsley Street by the Thames east of London (already mentioned in this chapter and in Chapter 2) is also extremely important here, because it indicates the existence of a contrasting tradition of inhumation burial in coffins at exactly the same time that repositories for the remains of multiple individuals began to be built in southern Britain, shortly after the start of the Neolithic period. Furthermore, the existence of cremation burials associated with fine carinated bowl pottery at Bishop’s Cannings in Wiltshire, at Yeavering in Northumberland, and at Midtown of Pitglassie in Aberdeenshire indicates that this diversity of practice was maintained through the early centuries of the Neolithic. What it suggests is that the deceased were in some cases afforded a burial that placed them outside the orbit of continuing daily life, while in other cases their remains began to be used as a kind of resource, possibly understood collectively as matter that was imbued with the presence of the ancestral dead, and which could be employed in a variety of activities that deliberately evoked the past.
When we turn to the more numerous long barrow and long cairn monuments that began to be constructed from 3800 bce onwards, our understanding of their individual histories has been transformed by the development of more fine-grained radiocarbon chronologies over the past decade or so. A detailed study of the construction-, use-, and decay-sequences of five long mounds in central southern Britain was the first archaeological excursion into Early Neolithic close chronologies. This series of studies found that the excavated one of two Ascott-under-Wychwood barrows north of the Thames in west Oxfordshire was most likely the earliest of this group of five closely studied monuments to have been built, and that its entire span of use was probably less than fifty years either side of around 3725 bce. Its construction may have been contemporary with the timber mortuary structure that was the first element in the development of the Fussell’s Lodge long barrow in Wiltshire. However, the elaboration of that structure and its subsequent transformation into a timber-revetted long mound was not completed for another century or more. It was not until as much as seventy-five years later than the start of building at Ascott that the primary elements of the Hazleton long cairn in the western Cotswolds were put in place. Another generation on saw the raising of the West Kennet long barrow near Avebury (but not its subsequent elaboration as seen today) around 3670–3635 cal bce (Fig. 3.5), while the first burials were placed in the chamber of the earlier of the two successive barrows at Wayland’s Smithy on the Ridgeway overlooking the Vale of the White Horse in Oxfordshire (formerly north Berkshire) at about 3610–3550 cal bce.
What is remarkable is that the entire sequence of interment in the first chamber at Wayland’s Smithy extended over a period of one to fifteen years. So the use of this structure for burial was brief, and limited to a few short episodes.
More recent research has demonstrated that the construction of the mortuary enclosure beneath the long mound of Wor Barrow on Cranborne Chase in north Dorset probably took place between 3685 and 3645 cal bce, much the same time as the construction of West Kennet. A turf mound was probably raised over the burials in 3650–3630 cal bce, and the ditch surrounding the mound was probably completed in 3630–3540 or 3585–3515 cal bce. The mortuary activity at this site is likely to have lasted for a little more than two generations. Aside from Wor Barrow, the refined dating of the initial five barrows is not yet closely replicable elsewhere because we do not have the same number of closely dated samples from other barrows: so the extent to which this sample of sites is representative of the wider picture is so far unclear. This is the case even elsewhere in the same regions as those barrows, although a further four sites in the Avebury area in which West Kennet is situated do provide some clues as to the likely variability among the twenty or so others known in and around the North Wiltshire Downs. These less detailed results indicate that the construction of barrows, some of them simpler in form, continued alongside the period in which causewayed enclosures were in use locally, which we discuss later in this chapter. Thus the long mound on Easton Down was probably raised between 3485 and 3385 cal bce, while that at South Street was almost precisely contemporary, with a likely build-date somewhere between 3495 and 3385 cal bce. An interesting feature of the latter site is that marks apparently made by an ard plough were scored into the pre-barrow turf sealed beneath this barrow, perhaps indicating that ploughing had taken place in two directions here not long before the barrow was raised. The stone-chambered long mound at Millbarrow appears to have been constructed late in the millennium, between 3390 and 3200 cal bce, while the barrow at Beckhampton Road west of Avebury, which may have contained no burial chamber at all, was built at much the same time. This tends to support the long-held conviction that long barrows built after the zenith of this kind of monument (which now appears to span a roughly 250-year period from around 3700 to 3450 bce) may have become increasingly divergent in form.
It has long been recognized that although earthen long barrows and long cairns generally contained multiple burials, these were far from numerous enough to represent the majority of the people who lived and died during the period. Given that the remains of men, women, and children are often represented, the criterion for selection was clearly neither age nor gender, and it has been speculated that these dead might have been drawn from a chiefly lineage, might represent persons who had died in inauspicious circumstances, or might simply be representatives of a larger community, deposited at infrequent intervals. The implication in each case is that aside from the few people who were cremated, buried in pits and flat graves, or deposited in caves, most of the deceased had been dealt with in some other manner that left no archaeological trace, whether exposed on platforms, consumed by scavengers, or immersed in rivers.
In both wooden and timber chambers, human remains have often been encountered in a state of disarticulation, with skulls often placed alongside each other and longbones bundled together. At Wor Barrow in Dorset, two bodies, at least one of them a young man, were laid out in a timber box chamber, before a third, partially articulated corpse was added. Later, three ‘bundled’ skeletons were added before the chamber was encased in turf. Later again, the bodies of a man and a child were introduced to the ditch surrounding the mound. Although the former of these was found in articular order, radiocarbon results indicate that he had died somewhere between 30 and 165 years before the ditch had been dug. Evidently, his remains had been curated, either tightly bundled or actually mummified. These findings highlight the importance that was sometimes afforded to remains regarded as ancestral, and indicate that not all the bodies interred in long barrows and long cairns need have been those of the recently deceased. People who were especially significant in relation to kin ties and descent may have been prioritized for inclusion in certain mortuary deposits at particular junctures. We will discuss the implications of these funerary practices further in Chapter 4.
Outside southern England the dating of long mounds and cairns is so far somewhat less precise. In Yorkshire, long barrows at Kemp Howe, Wold Newton, and Burythorpe were all probably built in the thirty-eighth or early thirty-seventh centuries, anything from a few decades to more than a century after the first Neolithic activity in the region. At Whitwell in northern Derbyshire, human remains probably began to be deposited in a linear mortuary structure in 3780–3720 cal bce, and this was later incorporated into a trapezoidal long cairn, following the construction and extension of a smaller oval cairn (see further discussion in Chapter 4). In the Scottish lowlands, long barrows may have first been built around 3800 bce, almost as soon as any other Neolithic activity has been identified locally. By contrast, Vicki Cummings has argued that many of the chambered tombs of western Scotland may have been no older than 3700 bce (Fig. 3.6).
This suggestion chimes with recent work by Rick Schulting and colleagues, which indicates that the court cairns (or chambered long mounds) of Ireland were first built in the mid thirty-seventh century bce, shortly after the climax of house-building, and more than a century after the beginning of the Neolithic there. This raises the possibility that in some areas mortuary monuments became common only after houses or halls began to go out of use. The implication is that these ‘houses of the dead’ took on some of the functions of the houses of the living, as a tangible feature of the landscape tied to the persistent identity of a particular human community.
It is now widely accepted that the mortuary practices manifested in long barrows and long cairns often took the form of protracted rites, in which fleshed bodies were laid out in a defined space such as a chamber and allowed to rot down, before the skeletal remains were subsequently reorganized. This has sometimes been interpreted as a means of expressing the person’s gradual incorporation into an ancestral community after death. However, it would be a mistake to extrapolate from these extended mortuary activities to imagine that the resulting deposits had accumulated over a period of several centuries. Instead, it has recently become apparent that the period of funerary activity was probably limited to a few generations at most. At Hazleton and at Ascott-under-Wychwood in the Cotswolds, for example, the refined radiocarbon chronology indicates that human remains were deposited over an interval of about a century. At West Kennet near Avebury in Wiltshire, the period of activity was limited to less than thirty years, and in the small earthen barrow that preceded the massive stone-bounded long mound at Wayland’s Smithy in Berkshire, it can have been no longer than fifteen years. Yet in most cases there are indications that the monuments concerned continued to be visited and venerated by their host communities over much longer periods. Animal remains or artefacts were introduced to the chambers or flanking ditches, and the mounds themselves were sometimes altered or enlarged, in various ways.
Perhaps the best way of explaining this is to relinquish the idea that long barrows and long cairns were the equivalents of modern cemeteries, that is, that they were spaces that were reserved for the accumulation of the dead over a lengthy period. Instead, inhumation was practised over a relatively short timespan integral to the construction and/or earliest use of each monument. So, rather than barrows and tombs serving as facilities for housing the deceased, it was the actual but in some senses token presence of the ancestral dead that conferred a particular social and historical status on these structures. In this way, it may be preferable to think of these works involving communal investment as commemorative structures, continually bringing to mind among the living their lineage stretching back to the foundational progenitor-groups: in other words, to significant ancestors. Long barrows and cairns embodied the community and its history in this way, because they contained the remains of that community’s founding generations. If ‘being Neolithic’ involved maintaining new mindsets, practices, and habits, as well as living in new ways alongside plants and animals, these were the people who had first dwelt in this manner. They constituted a point of origin for their descendants, who both reckoned their descent and drew their identity from these ancestral generations.
Often, nonetheless, the construction of long mounds containing chamber spaces for the dead followed on from earlier activity in the same location, which may not have been strictly funerary in character. In many earthen long barrows, the chamber took the form of a linear, trough-like space flanked by banks of clay or stone and bracketed by two massive upright posts (sometimes with a third, medial post). These posts were generally derived from the trunk of an oak tree, which had been split in two longitudinally, with the two flat sides facing inwards. In some cases, as at Street House in North Yorkshire, and at Haddenham in the Cambridgeshire fens, these timber uprights are now thought to have been decades or even centuries older than the rest of the chamber to which they later became integral. The implication is that pairs of posts originally constituted freestanding structures of some kind, and that their funerary use was a later and secondary adaptation. Interestingly in this context, pairs of oak uprights often form an element of the large timber halls that we have already discussed, and it may be that these pairs of posts were regarded as timber shrines which connected the ancestral community with the subsequent creation of structures associated with mortuary and commemorative rites. The raising of posts bedded in the earth in this way may have made a powerful foundational statement that was both sustained and at the same time reinforced by the transposition of the posts and their symbolic enclosure of human remains.
Regionalization and exchange networks
If the early part of the fourth millennium saw the establishment of a Neolithic way of life in Britain, aspects of this came to be transformed in fundamental ways during the thirty-seventh century bce. One indication of the nature of such change is registered in pottery production and use. Over time, the fine carinated bowl pottery of the earliest Neolithic occurred within more varied and larger groups of pottery vessels, became more regionally diverse, and may have begun to be used for a wider variety of purposes (Fig. 3.7).
The ‘developed carinated bowl’ at the end of this series is notably thicker and coarser, and manifests in a wider range of vessel forms. Meanwhile, in the south and west of England assemblages occurred that combined coarser carinated bowls with other forms, apparently drawing on west French prototypes. Subsequently, throughout much of Britain, ‘heavy bowl’ assemblages developed, with a variety of hemispherical bag-shaped and shouldered forms, thicker walls, and heavy rims. It is hard to avoid the conclusion that pottery had by this time lost something of its exotic and exclusive character, although its increasing ubiquity occurred at the same time as the emergence of a series of decorated wares. These latter were often no finer than the plain pottery, and assemblages frequently combine the two, in proportions that vary from region to region. However, what is notable is that one particular vessel form, the shouldered bowl, was repeatedly singled out for decoration, and this was generally concentrated on the rim and the upper part of the body. Joshua Pollard has recently pointed out that the burnished surfaces and subtly rippled decoration of the earliest pottery suggests that it was engaged with in a tactile and intimate manner, perhaps by small groups of people. In contrast, the introduction of decoration indicates that pots may have increasingly been apprehended visually, amidst larger gatherings, and their significance appreciated from a distance.
It is with the development of decorated pottery that distinct regional identities begin to be clearly discernible in Neolithic Britain. It may be that groups of people living in different parts of the country had gradually and unintentionally developed separate traditions of ceramic manufacture, in the manner of what has been termed by anthropologists ‘cultural drift’. However, studies of near-contemporary but essentially pre-industrial technology demonstrate that people are generally acutely aware of the differences between their own handiwork and objects made by neighbouring groups. Moreover, they actively use these differences to establish and maintain boundaries of one kind or another. However, these ‘decorated wares’ actually tend to blur into each other to some extent, and their ornamentation may have had as much of a role in marking out places and practices as separate groups of people. In most cases, the new mid fourth-millennium decorated styles of Neolithic pottery in Britain clearly developed from, but involved an elaboration of, existing vessel forms. For example, the Whitehawk Ware of south-east England featured open, S-profiled, round-bottomed bowls closely related to the earlier fine carinated bowl assemblages, but with decoration on the upper body above a horizontal cordon. Meanwhile, the Windmill Hill Ware of northern Wessex included among its variety of forms decorated versions of shouldered bowls apparently inspired by pots that were in use at broadly the same time in western France.
This proliferation of ceramic traditions and growth of decoration coincided with the occurrence of a series of indicators of social stress and enhanced competition. One of these was a growing number of human bodies showing signs of either healed or fatal trauma. These include the tips of flint arrowheads embedded in human ribs, and depressed fractures in skulls, suggesting the impact of a blunt weapon. Observations of these traces of body traumas are now numerous enough and are sufficiently closely dated as to indicate the existence of endemic warfare, although the probability is that this generally took the form of raiding and feuding, and that pitched battles were infrequent events (see Chapter 4 for further commentary). This picture echoes the evidence for a number of massacres earlier in the Neolithic era in continental Europe, and indicates that the period more generally was not necessarily one of widespread peace and cooperation. Rather, the periodic threat of violent interactions formed part of the background to everyday life in the Early Neolithic. This was nonetheless the counterpart to practices that increasingly emphasized the importance of alliances, collective hospitality, and convivial relations. We might therefore suggest that social life was becoming more intensive, in various ways and at several different scales.
This accords well with the way that the growth of regional identities from the thirty-seventh century bce onwards is matched by increasing evidence for interregional contact and exchange. Such is the range of artefacts and materials that was being circulated over considerable distances from this time onwards that it can surely be inferred that a set of more extensive exchange networks was in the process of being established. Good-quality flint from Beer Head in Devon, and dark chert from the Isle of Portland in Dorset, for example, were transported into central southern Britain, as were pottery vessels gritted with fragments of igneous gabbroic rock from the Lizard peninsula in Cornwall, and with oolitic limestone from the area between Bath and Frome. More widely distributed still were ground and polished stone axeheads from quarry sites in the west of Britain. These might arguably have emulated the axes of jadeitite that had been quarried in the Italian Alps from the sixth millennium bce, small numbers of which had found their way to Britain. It is likely that one of the earliest of the British stone sources to be exploited was the distinctive blue-green volcanic tuff of the Langdale valley in Cumbria, where radiocarbon dates for the working are concentrated in the period between 3800 and 3500 cal bce, but continuing sporadically for a further five hundred years thereafter. The appearance of the Langdale axeheads was not entirely unlike those made in jadeitite stone from Alpine quarries, and similarly the Langdale stone itself was found, as with the jadeitite source-locations in the Alps, in a relatively inaccessible upland location. Moreover, the most common source-location was distinguished by its occurrence close to or on the visually arresting Langdale Pikes.
Fieldwork in the Langdales by Richard Bradley and Mark Edmonds has demonstrated that the earliest quarrying for axes may have been relatively unskilled, resulting in numerous failed axe preforms, and waste flakes of variable size and thickness. Over time, it seems that the manufacture of axes became more accomplished, and was concentrated in a smaller number of locations, which tended to be more difficult to access, and to command more spectacular views. This is most clearly the case with the quarrying on the buttresses of the Pike of Stickle, where fires were lit to loosen the rock from the quarry face (Fig. 3.8).
It is probable that while the earliest stone axes from Langdale were made by non-specialists who visited the valley seasonally, perhaps grazing animals on the slopes and in the nearby valleys, over time more skilled people took over, presumably making axes intended for long-distance exchange. Indeed, the densest concentration of Langdale axes found at some distance from their source-area is located in the Wolds of East Yorkshire, where there is abundant flint that could be, and was, used to make perfectly serviceable polished axes. The implication is that axes which came from a remote and special location, and which had a distinctive appearance, had a value over and above any everyday utility they may have had for felling trees, splitting logs, or clearing brush.
Valuable objects acquired from a distance and limited in their supply might have had a number of important roles. Secured in a wooden haft and carried around, a finely polished stone axe would have announced its owner’s importance or identity, and drawn attention to their access to connections with distant communities and places. As a gift, it would have enhanced the giver’s prestige, attracted followers, or secured alliances. As a payment, it might have procured a marriage or compensated for a killing. It is therefore easy to see the attraction that goods of this kind may have had in the increasingly violent and competitive world of the earlier fourth millennium. Indeed, the fine but nonetheless strong stone axeheads that have in rare circumstances been found along with their elaborate wooden hafts may themselves have been weapons of both display and actual warfare.
In time, the axes from the Lake District were complemented by those quarried at, and procured from, stone sources elsewhere in Britain, for instance in Cornwall, north and south-west Wales, and western Scotland. Even axes of the distinctive dark porcellanite stone from Tievebulliagh in Antrim and from Rathlin Island off the north-west coast of Ireland sometimes occur on the British mainland. What all these source-locations share in common is remoteness, and frequently an association with dramatic crags or mountainous outcrops, several among which directly overlooked the sea.
This rapid ‘uplift’ in the tempo of the interregional exchange of readily identifiable stone items was accompanied in all likelihood (albeit as gauged from so far only exploratory studies) by the use of certain rocks, after deliberate crushing, as tempering agents in pots, including the gabbroic and oolitic sources noted already. That these pots themselves had in some cases travelled some distance from their location of preparation or production tells us something about the nature and purpose of the supra-community interaction that was occurring at this time. So although regional identities had become more distinct by the middle of the fourth millennium, they were cross-cut by channels of exchange contacts that extended over very considerable distances. This may have resulted from, but may also have partially facilitated and fuelled, more localized competition for personal or kin-group-based accumulation of wealth, prestige, and the acquisition of followers.
A narrative of changing subsistence practices
The development of new ways of circulating desirable and exotic artefacts and materials was complemented, perhaps not entirely surprisingly, by changes in the way that food was acquired. We have seen that at the beginning of the Neolithic domesticated plants and animals were introduced to Britain from the Continent. To begin with, the cultivation of cereals was seemingly adopted enthusiastically throughout much of Britain, although probably in very small ‘garden’ plots, and alongside the continued exploitation of wild plants such as hazelnuts, acorns, berries, and crab apples. However, from the thirty-seventh century onwards the representation of directly dated charred cereal grains and threshing residues in archaeological deposits on the British mainland begins to fall away abruptly, becoming all but invisible by 3300 bce. There is strong circumstantial evidence that cultivation did not cease altogether, but its scale and intensity almost certainly declined. The remains of wild plants, however, continue to be present among the dated samples that have been recovered from this time horizon. A similar pattern has been observed in Ireland, and here it has been noted that the apparent decline in cultivation coincides with the disappearance of the numerous rectangular timber houses that, as we have seen, were built during a short period in the earliest Neolithic. One interpretation that has been put forward for this phenomenon is a ‘boom and bust’ of pioneer agriculture. It has been noted that the yields achieved from early, unimproved grain species were generally low, but that for a short while after their introduction into a given area they would have been relatively productive. However, after this ‘honeymoon’ period, local pests and diseases would have begun to attack the new domesticates, while soil exhaustion may also have started to set in. The consequence of this will have been a collapse of arable farming, and a rapid decline of population, which only recovered in the Later Neolithic.
The flaw in this argument is that the period between, say, 3650 and 3300 bce actually coincided with a major increase in the scale and number of the monuments being built in both Britain and Ireland. In Ireland, the supposed agricultural crash corresponds with the development of the passage- tomb tradition. On the basis of a project devoted to dating the bone pins from the Carrowmore passage-tomb cemetery in Co. Sligo, Robert Hensey has argued that small, simple tombs with polygonal chambers, surrounded by a boulder kerb, may date to the thirty-eighth or thirty-seventh century bce, shortly after the start of local Neolithic activity. After 3600 bce, much more numerous and more substantial structures developed, like those in the Loughcrew and Carrowkeel cemeteries, with substantial covering mounds and more developed passages and entrances. The process of elaboration in scale and complexity culminated with the construction of the three massive tombs of Knowth, Dowth, and Newgrange, in the ‘bend in the Boyne’ in Co. Meath, between 3200 and 3000 bce. This development, demanding the investment of progressively greater quantities of labour, was played out over the period during which the supposed economic decline took place.
In southern Britain, the same period saw the creation of two of the country’s most monumental prehistoric structures, the Dorset Cursus and the Hambledon Hill causewayed enclosure complex, each of which is discussed further in this book. If large numbers of people could be mobilized to work on the creation of these edifices, the notion of a catastrophic fall in population is probably not tenable. Equally, while the evidence for cultivated plants begins to fall away at this time, the same is not the case for domesticated animals. This evidence suggests, on the contrary, that there was an increasing emphasis on raising cattle: it has been argued that at some enclosure sites the animal bone assemblages indicate the culling of older cows from large herds that were not concentrated in the immediate vicinity year-round. Chemical analysis of residues from pottery from the start of the British Neolithic onwards has not only shown that it is highly likely that dairy products were regularly being consumed, possibly including cheese as well as milk, but also that this consumption seemingly continued after the thirty-seventh century bce. Moreover, strontium and oxygen isotope values from the tooth enamel of people buried in this period are such as to suggest that they had often repeatedly moved during their youth between areas some distance apart, and with different geological characteristics. New analyses of bodies dating to the thirty-seventh century bce from Wor Barrow in Dorset, for instance, suggest seasonal movements (or perhaps resource acquisition) between different regional environments. Comparable work on remains from Hazleton in the Gloucestershire Cotswolds indicates more extensive peregrinations, as far as Herefordshire or south Wales.
One plausible way to explain all this evidence is as representing a change of emphasis from arable to pastoral farming, with at least a proportion of the population following large mobile herds between seasonal pastures. Probably both wild and domesticated plants continued to be exploited, although there are hints that the plots where the latter were grown were less intensively manured as time progressed. Nonetheless, the centre of gravity of subsistence activity seems to have shifted markedly towards meat and dairy products. It may be that a decline in cereal yields provided a ‘push’ factor behind this change, but the ‘pull’ of livestock was probably the more important factor in promoting the shift.
A key reason for this is, we think, that while plant foods are reliable staples that feed people and keep them alive, livestock represent mobile wealth as well as simply nutrition. Herds of cattle, in particular, can be accumulated by prominent people or powerful groups, and constitute ‘capital’, the use-value of which can be realized in a variety of different ways. While milk and a certain amount of meat can contribute to the everyday diet, meat is much more valuable than grain as a feasting food, and the provision of feasts is a powerful means of enhancing prestige and social standing and recruiting followers. Cattle are also better suited than cereals for use as dowry or bridewealth payments, and as gifts they can establish relations of debt and obligation between donor and recipient. By building up increasing numbers of cattle, it would have been possible for particular people to actively manage and steer their social interrelations in a variety of ways, thereby increasing their influence and authority. It is probably no accident, in this light, that the start of this shift from cultivation to herding appears to have coincided with the growing evidence for regionalization, violence, and the exchange of prestige valuables. This is because all these developments denote the emergence of a more competitive society and increasing social inequality. Indeed, it is entirely possible that some at least of the violent episodes noted earlier could have occurred during cattle-raiding expeditions.
However, in one part of Britain a very different pattern of settlement activity was developing at this time. In the Orkney Islands, Neolithic activity began a little later than on the mainland, and a decline in cereal cultivation from the thirty-seventh century is not evident at all. While the large timber halls and smaller houses were a short-lived phenomenon elsewhere in Britain, there is a virtually unbroken sequence of the building of walled and roofed structures throughout the Orcadian Neolithic, beginning with scattered post-hole-built timber houses and developing into nucleated villages of cellular stone buildings. This was not simply a consequence of the existence of suitable building materials on the islands, since areas such as Caithness that have precisely the same flagstone and sandstone geology have (so far) not revealed any settlements of Neolithic stone houses. While much of Britain moved towards a semi-mobile way of life, based on the accumulation of cattle herds, this may not have been practicable on the islands, where ecologically distinct and spatially separate seasonal pastures were not available, and stands of hazel were minimal. In its place, a distinctive social and economic formation developed, based on household groups with fixed horticultural plots and small numbers of livestock. The emergence of this very distinctive Orcadian framework of economic and social life was, remarkably, and as we shall see, to have far-reaching consequences across Britain later in the Neolithic.
Building long mounds and chambers, variously
One of the problems highlighted by the development of a more accurate chronology for the Neolithic is that although individual mortuary monuments may have been constructed and used over a relatively short period of time, broadly similar kinds of structure may have been built over several centuries. It follows that neat sequences in which one morphologically defined ‘type’ is replaced by another are often less accurate than previous generations of archaeologists would have assumed.
Indeed, the very structural features of long barrows that were once thought to have indicated that they were built and used at the same time as one another may instead denote the recurrence of motifs and practices that had their heyday and then became part of a repertoire that enabled them to be redeployed in distinctly later monuments. Only the kind of fine-grained chronologies already mentioned will enable us gradually to unpick these complexities. Some broad patterns are nonetheless emerging. We have seen that long mounds and cairns containing stone chambers first occurred quite early in the British Neolithic sequence. However, their appearance in different regions was not precisely synchronized, nor was their occurrence prefigured by a uniform elapse of time after the first manifestation of Neolithic things and practices locally. In general, there appears to have been escalation of monument building from the thirty-seventh century onwards, which involved both a proliferation of existing kinds of structure and the emergence of new traditions. Long mounds and cairns in areas such as southern England, eastern Scotland, and eastern Yorkshire continued to be constructed until after 3500 bce, although, as we have seen, the primary use of any particular barrow for the deposition of the dead often lasted for less than a century.
Yet there is some indication that as the long-mound tradition progressed, it became more diverse. Some mounds that have produced relatively late radiocarbon dates, such as the Amesbury 42 long barrow located beside the Great Stonehenge Cursus, are exceptionally large. Other massive long barrows, such as Tilshead Old Ditch on Salisbury Plain, may have encapsulated smaller, earlier long mounds.
In addition, prodigiously long mounds known as ‘bank barrows’ developed, such as the potentially two-kilometre long Tom’s Knowe/Lamb Knowe structure that crosses the White Esk in Dumfries and Galloway. At Maiden Castle in Dorset another bank barrow runs across the banks and ditches of the Neolithic enclosure, demonstrating that it dates to a mature phase of the Neolithic. Bank barrows are of consistent height and width along their length, and contain no obvious mortuary structure or chamber. Like the equally late South Street and Beckhampton Road long barrows in the Avebury area that we have already mentioned, they rarely produce any human burials. Towards the end of the long-barrow tradition, then, mounds appear to have been built whose sheer imposing monumentality was more significant than their potential role as containers for (some or all of) the dead of each relevant community. Other late long mounds, of which the Alfriston long barrow in East Sussex is a good example, contained small numbers of articulated burials in individual graves or small-group interments, rather than featuring any kind of collective mortuary structure within the mound. This suggests a form of funerary activity of yet shorter, rather than longer, duration, and a mortuary ritual focused on maintaining the bodily remains intact rather than the earlier practice of separating flesh and bone, and then disaggregating the separate elements of the skeleton, which were often reinterred elsewhere.
As we saw earlier in this chapter, it was not uncommon for individual barrows to have a complex developmental history, with significant changes being registered in the form of the structure, or even its wholesale replacement. In the case of the five sites subject to the first detailed Bayesian chronological studies, the mound at Ascott appears to have been enlarged less than fifty-five years after its initial construction, while Hazleton was constructed as a single, if perhaps protracted, episode. The chambers and mound at West Kennet probably represented a single event of building, but kinks observed in the side ditches may indicate that the tail of the mound was an addition to an initially smaller tumulus containing the chambers, resulting in a truly colossal structure. At Fussell’s Lodge the wooden mortuary structure was in use, and enlarged, over a period of more than a century before the massive mound was raised over the ‘bone repository’, bringing its use to an end. The similar wooden chamber at Wayland’s Smithy was in use for a much shorter period before it was sealed beneath a small earthen mound. Around a century after this first chamber was built a more dramatic change took place: an entirely new structure was built over and encapsulated the earlier mound, on a monumental scale. Not only was this structure at Wayland’s Smithy massive and built with a stone kerb defining its edges in a huge trapezoidal shape, but a façade comprising massive individual sarsen stones was built to create a forecourt with, as its central focus, a stone-lined and walled passage that extended into the new mound to give access to a whole group of chambers radiating from it (Fig. 3.11).
The chambers were themselves lined and roofed by massive stones in a closely similar way to those at West Kennet and other ‘chambered tombs’, and they then became the receptacles for the interment of a further series of mortuary deposits. While the earlier burials were placed largely intact upon a ‘pavement’ of flat stones within the first barrow, where they slowly collapsed and became intermingled, the bones placed in the stone-lined chambers at the subsequent, more massive Wayland’s Smithy barrow would have been accessible despite temporary blockings. The term ‘repository’ is therefore all the more apt for this later structure (Fig. 3.12). One question that inevitably arises from appreciation of this sequence and complexity is how typical it was of the hundreds of other known sites of long barrows not only in southern but also in western and northern Britain. As yet, unfortunately, there are too few sets of dates to be able to create detailed chronologies of the kind already mentioned: but it is clear that in time some fascinating patterns of co-occurrence and development are likely to emerge to amplify further this key aspect of Neolithic history.
The greater diversity of barrows that began to be created in western and northern Britain after around 3700 bce often formed distinctive regional groups. It has been argued that some simple passage tombs, distinguished by a box-like chamber of stone uprights accessed by a short, walled passage in a small round cairn, may be earlier than this, even perhaps dating to the fifth millennium bce. However, there is at present little unambiguous evidence for this. We have seen that the example at Broadsands in Devon dates to the thirty-eighth century bce, and that the new evidence from Carrowmore suggests a similar or slightly later date. Nonetheless, a case has been made that the excavated cremations and their accompanying antler pins may post-date the construction of the tombs, and that earlier deposits may have been removed, leaving no trace. Recent radiocarbon analysis from the passage tomb of Baltinglass Hill in Co. Wicklow has produced one result of 3946–3715 cal bce from a calcined human cranial fragment, but others that fit more comfortably in the thirty-seventh century. The single, anomalously early date might result from the introduction of curated bone into the tomb, or might simply be inaccurate, placing the site temporally alongside the new dates for Carrowmore. A more radical view has it that many of the Carrowmore monuments are not passage tombs at all (but simple dolmens that were later architecturally elaborated), that Broadsands is too ruinous for its affinities to be realistically assessed, and that the earliest passage tombs in Britain and Ireland are actually the Maes Howe-style monuments of Orkney, such as Quanterness. The Baltinglass results perhaps argue against this perspective. We have already noted that long cairns in Ireland and western Scotland began to be built in appreciable numbers in the thirty-seventh century bce, but there may have been earlier outliers within this overall pattern. An example might be the recently excavated ‘Clyde’ cairn at Blasthill on Kintyre, which was possibly built at an early point in this sequence.
The builders of most of these groups of barrows ultimately took their inspiration, and borrowed particular constructional devices, from the Continent, especially western and northern France. However, this does not mean that they had always seen the precursor monuments for themselves: the stylistic connections are generally imprecise, and architectural elements from different sources (potentially including southern Scandinavia and the North European Plain) were seemingly mixed together and elaborated upon. Meanwhile, the distinctive ‘portal dolmens’ of Wales, Ireland, and Cornwall have no precise Continental parallels at all (Fig. 3.13). These consist of a massive capstone raised on a series of supporting uprights, including an arrangement of three stones that might have mimicked a doorway. The effect of this would have been that the chamber was closed, so that continued access to any mortuary deposit was presumably not a priority. Indeed, it may be that the display of the capstone was the principal objective in constructing the monument, and that the incorporation of human remains may have been a secondary issue. This could be compared with the pairs of earthfast upright posts that we have seen were often incorporated into timber mortuary structures, but that might initially have been more significant as a conspicuous arrangement of massive timbers. Portal dolmens apparently had small or vestigial mounds or cairns, so that the large capstone would probably always have been visible. While all the barrows that included large stones in their make-up may have contained the remains of the dead (although this cannot be demonstrated in many areas with acidic soils, where bones have not survived), making a visual impact on the landscape sometimes appears to have been of equal or greater importance (Fig. 3.14).
As for the date of the appearance of this distinctive form of megalithic structure, there remain uncertainties. At Dyffryn Ardudwy in Gwynedd, for example, a small pit containing the sherds of four carinated bowls was found immediately in front of the portal stones of such a structure, suggesting an early beginning, perhaps as early as 3800–3700 bce. This impression is reinforced by the fact that the structure was later incorporated into a long cairn that featured the piling up of stones around and perhaps over the original dolmen. If typical, this would imply a date early in the Neolithic for portal dolmens, but the radiocarbon evidence presently remains equivocal. Dates on human remains from the Poulnabrone dolmen in western Ireland are also early, but their relationship to the chamber is unclear. In the meantime, if we are to judge from the existing available radiocarbon dates, many of these dolmens may date instead to the mid fourth millennium bce.
An appreciable proportion of these monuments underwent multi-phase construction like that at Dyffryn Ardudwy (Fig. 3.15). That is to say, an initial structure that may have been relatively diminutive had a series of additional bodies of cairn material added to it. In some cases this simply rendered the site much more massive, but in others the addition of a tail, or long cairn, gave it a distinct orientation within the landscape. It may be that this practice was simply a matter of growing communities being able progressively to render up greater amounts of labour to honour their ancestral dead. But equally, it may instead signal a shift of priorities, towards the creation of a more visible imprint on the land, and a new specification of the conditions under which the living could approach and experience the resting places of ancestral remains.
Creating enclosures and delimiting space within a circle or curve
The increasingly common creation of barrows featuring large stones in the west of Britain coincided with the introduction of a new kind of monument, initially in the south-east of England. Causewayed enclosures are so termed because they feature one or more (mostly) concentrically ordered rings of interrupted ditches, ranging between 100 and 600 metres in diameter (Fig. 3.16).
Some of these rings form complete circles or oval-shaped enclosures; some sites feature multiple, sometimes linked rings; and others form arcs of ditch that cut off promontories or are sited next to rivers or steep scarps. Around eighty of these sites have been recognized in southern Britain, although less than half have featured excavation by modern standards, and many have not been examined at all. Although the sausage-shaped segments of ditch may be associated with a bank or palisade, and the bank may be continuous, this is not a means of construction that in any sense maximizes the defensive potential of any of the sites in question. Ditched enclosures developed on the Continent, in earlier stages of the European Neolithic. Small enclosures were a feature of the later decades of the Linearbandkeramik villages of Germany and Belgium, while larger sites developed during the later fifth millennium bce. Enclosures with interrupted ditches, banks, and palisades were a feature of Chasséen and Michelsberg communities in Germany, northern France, and the Low Countries. Examples very similar to the British ones are found in the Rhine valley and the Pas-de-Calais. Yet from the earliest, the British enclosures were associated with styles of pottery and stone tools that had developed in the insular context (and coincided also with the emergence of the decorated ceramic styles discussed already). They rarely contain any kind of internal structure other than groups of pits, and their use has been a topic of debate for many years.
In the Gathering Time study, some twenty-seven among the seventy-four sites recognized as causewayed enclosures (by the period of data collection between 2006 and 2009) were found to be amenable to detailed chronological evaluation. One of the most significant facts underlined by the study was that nearly all the earliest-dated enclosures occur in south-eastern England. Moreover, the earliest dates, going back to around 3750 bce, have been obtained for sites overlooking the coastal plain of Sussex and located on and around the Thames Estuary. If this chronology is an accurate representation of reality, it is difficult to avoid the conclusion that it was in precisely those areas that were in a position to maintain close contact with the Continent that this innovation of enclosure was first developed. Given the increasing contrasts in material culture that by this time existed between the Continent and Britain, it is likely that the idea of enclosure may have travelled with individuals rather than representing a mass migration as such, and in these terms causewayed-enclosure creation may have involved a deliberately adoptive process.
Clues as to the role of causewayed enclosures in the Neolithic period in Britain were identified early on in their archaeological investigation. One of the first British sites to be examined was Windmill Hill near Avebury in Wiltshire, excavated by Alexander Keiller in 1925–9. The results were eventually published by Isobel Smith four decades later, and on the basis of the large quantities of non-local pottery and stone axes recovered from the ditches, she proposed that the enclosures had served as meeting places and marketing-centres for dispersed populations. More recent excavations at the same site conducted by Alasdair Whittle and Joshua Pollard have demonstrated that very large numbers of animal bones, predominantly cattle, had been deposited in the ditches. This would suggest that the collective consumption of meat in appreciable quantities, arguably during feasting events, also occurred at some of these locations. At a number of sites it has been inferred that the cattle remains represent only a fragment of very large, mobile herds, predominantly either young adult females, or older cows culled when they were beyond milking age.
At another extensively investigated site, on Hambledon Hill in Dorset, two separate enclosures and two long barrows were enclosed within a massive system of outwork banks and ditches covering a series of interconnected chalk spurs (Fig. 3.17).
This complex developed piecemeal, over a period of time, beginning at 3680–3630 cal bce, with activity on the hilltop continuing over a period of 310–70 years in total. The larger, main enclosure was one of the earliest elements, and contained numerous pits filled with deer antler, pottery fragments, sandstone rubbers, and stone axes, while its ditches held appreciable quantities of human bone. These included skulls that had clearly been placed on the base of the ditch. One possibility is that the enclosure had been used as a ‘cemetery’, where the dead were exposed to the elements, with fragments of bodies entering the ditch as a consequence of the activities of carrion eaters. However, in some cases the remains had apparently been deliberately placed, while the presence of marks caused by scraping the flesh from bones with stone tools indicates that corpses were sometimes actively ‘processed’ rather than simply left to decay (see Chapter 4 for further discussion of mortuary practice at Hambledon Hill).
Isotopic analysis of these bones has shown that the people concerned had enjoyed a greater diversity of diets than would be represented amongst the burials in any single tomb or barrow of equivalent date. The implication is that the people concerned belonged to a number of different communities, and came from a relatively wide geographical area. The second enclosure, on the Stepleton spur, had been constructed a few decades after the main site and may have been sporadically occupied rather than used for ceremonial activity. The system of cross-ditches and outworks became progressively more elaborate, and it is implied that this activity was a response to a growing threat of violence. Two episodes of aggression appear to have taken place in the mid fourth millennium bce. In the first, the inner Stepleton outwork had been burned, shortly after its construction, and a mature adult male was found on the base of the ditch. Eighty metres away, a young man had been buried with a mass of burnt material. Perhaps a century later, the body of a young man who had died by arrowshot entered the ditch of the outer Stepleton outwork, while another was found on top of the rubble fill of the inner Stepleton outwork ditch. Broadly contemporary with these two violent deaths was the articulated upper body of a woman, found in the ditch of the Stepleton enclosure, which was now largely silted.
Similar traces of burning and the shooting of large numbers of arrows were found at the enclosure on Crickley Hill in Gloucestershire, which stands out from the scarp overlooking the Severn Valley near Cheltenham. The early dates for the Ascott and Hazleton barrows nearby are complemented by slightly later dates for the earliest of the arcs of ditch that cut across the promontory at Crickley. This location had already seen human activity in the form of a series of small, circular stake-built structures, which may represent a Mesolithic occupation, and a group of small pits enclosing an irregular eight-metre long oval, referred to as the ‘Banana barrow’. The outer and inner causewayed ditch circuits of modest size were dug around 3660–3610 bce, and these had been deliberately backfilled and re-dug on numerous occasions. Only toward the end of the use of the site had a much more substantial, continuous, and defensive ditch been dug, in the period between 3565 and 3535 bce, complemented by a stone rampart. The palisade at the back of this rampart had been burned, probably sometime after 3490–3450 cal bce, and numerous leaf-shaped arrow points were distributed along its length, as well as being concentrated in the entrances through the defences. Structures within the enclosure appear to have been burned at much the same time, and this episode may have brought activity on the hilltop to an end.
There is considerable variation in the period over which individual enclosures remained in use. Some may have been abandoned shortly after construction, but the enclosures on Hambledon Hill and Crickley Hill demonstrate that others endured for many generations. Where this is the case, enclosures developed complex biographies, in which structural alterations and changes in the character of activity would presumably have been matched by changes in meaning. The Hambledon and Crickley enclosures both became notably more defensive in character in time, but this was by no means a universal pattern. However, similar developments may also be demonstrated by the enclosure set around a prominent rocky granite hilltop ‘tor’ at Carn Brea near Redruth in Cornwall, and at the causewayed enclosure (also apparently set alight) on the inland hill-promontory at Hembury, north-east of Exeter in Devon.
The pattern noted at Crickley Hill, of causewayed enclosure ditches being repeatedly re-dug, sometimes after deliberate backfilling, is one that has been reported at numerous other sites. If the significance of enclosing an area with a discontinuous ditch is that it was primarily a means of defining a particular space as special, and appropriate for specific kinds of activity, this recutting would serve to renew and reiterate that status. In some cases, it might have been the means of re-establishing the temporary conditions required for gatherings, feasts, and inter-community negotiations or exchanges. A particularly spectacular example was discovered at Briar Hill near Northampton, where meticulous excavation and recording revealed an enormously complex series of recutting events (Fig. 3.18).
At Etton in the Cambridgeshire fens, a small, single-ring causewayed enclosure had been utilized for the best part of four centuries, although the density of finds in the ditches was modest by comparison to Hambledon or Windmill Hill. Having been subject to seasonal freshwater flooding, the site was waterlogged, and organic material had survived in the ditches. The eastern arc of ditches had been backfilled shortly after it was first dug, but it was recut on numerous occasions. There were many examples of carefully placed deposits. These included a complete bowl on a birch-bark mat, an intact axe-haft, and a bundle of cattle bones, as well as a series of animal skulls. At Etton, there appears to have been a close relationship between the practice of recutting and the placement of objects in the ditch segments.
There seems to have been considerable variation in the activities that took place at different causewayed enclosures. In the Neolithic enclosure that underlies the massive Iron Age hillfort at Maiden Castle in Dorset, it appears that flint axes were being finished, ready for use. By contrast, at Etton some of the Langdale stone axes represented had been deliberately smashed in the ditches. While the creation of interrupted ditch circuits transformed the status of the enclosed space, it nonetheless maintained its accessibility (although see Chapter 6 for a reading of the reason for segmentation that focuses instead upon a potential lineage identity and descent dimension to the practice). The transitory or temporary character of some of these sites (at least initially) that were repeatedly brought into being, and then ‘decommissioned’, implies that their use was sporadic and episodic, perhaps seasonal, as Stuart Piggott once argued. People, animals, and artefacts all came to these enclosures from elsewhere, often from great distances, and in each case significant events in their personal histories took place within the bounded space of the site. Artefacts were created, destroyed, and displayed, cattle were slaughtered, the recently dead were ‘processed’ in various ways, and food was consumed in large quantities.
It is reasonable to assume that other, less tangible transitions and transformations might also have taken place at causewayed enclosures: marriages arranged and celebrated, alliances negotiated, liaisons and friendships established, exchanges of valuable items conducted, disputes settled. All of this makes a great deal of sense in the context of the kind of society that was emerging in the second quarter of the fourth millennium bce: increasingly competitive and violent, less reliant on cultivation, more focused on the accumulation of herds and fine artefacts, more mobile and dispersed, more unequal. The establishment of temporarily occupied, politically neutral spaces where different groups could come together at particular times of year created the conditions under which the recurrent enmities of the time could be set aside, and the important business of exchanging goods, information, and marriage partners could be carried out. They could perhaps have been perceived as places of armed truce, but in some cases they may gradually have become associated with the activities of powerful social factions that were emerging at this time, or disputed and fought over by rival communities (Fig. 3.19).
The Gathering Time studies of Alasdair Whittle, Frances Healy, and Alex Bayliss sketched out a chronological narrative of changing patterns of use and development of these enclosures on a region-by-region basis, at least where numbers of them have been investigated in southern and eastern Britain, and where, especially within those regions, their sequences have begun to be closely dated. For example, by combining a variety of estimates of effort expended in digging the ditches of causewayed enclosures with the dating of different circuits of these enclosures, a fascinating picture has emerged of how and where most effort was put into their construction. Broadly, it looks as if the proportion of ditch length dug, mapped against the passage of time, rose rapidly from the second quarter of the thirty-seventh century bce, then fell markedly from around 3625 cal bce, reaching its lowest point at about 3575 cal bce. Construction activity had then picked up somewhat by 3450 cal bce. The early ‘peak’ of construction was apparently mostly a south coast and Thames-side/Cotswolds phenomenon, with the ‘end’ of the peak coinciding with the creation of enclosures in eastern and south-western England. Such a broad-brush portrait, of course, masks many complexities, and some of the detail concerns the timing of the digging of multiple circuits of ditches where these occurred, episodes of recutting of ditches, and the relative timing of construction of ‘pairs’ of sites.
We have seen that the length of time that such sites were in use also varied, and that several seem to have been short-term projects that began and ended within a single generation. The tempo of use is also an important consideration, and here the Gathering Time team introduced further innovation into their calculations. Estimates were made from known lengths of ditch from which quantified amounts of early ‘Neolithic Bowl’ pottery and struck flint or chert pieces were recovered, and comparisons were made between sites. This indicated that some of the shortest-lived enclosures such as Abingdon by the Thames and Maiden Castle near Dorchester in Dorset also featured the greatest intensities of deposition of artefacts in their ditch-segments. At some sites such as Hambledon Hill, at Windmill Hill, and at Etton, the initial digging of the ditches was accompanied by relatively few artefact deposits, while the later fills of the same ditches were often relatively artefact-rich.
Causewayed and related enclosures had a total primary (that is, Early Neolithic) use-span of between 400 and 450 years, but again, their falling out of use was neither gradual nor unimodal. There was a period of little more than fifty years either side of 3500 bce when over 30 per cent of the total went out of use. By approximately 3400 bce, however, the number falling out of use, compared to the (by then) total number of sites, had reduced rapidly to less than 25 per cent, largely owing to the commissioning of new sites or the creation of new circuits, or the recutting of previously dug segments. This number of ‘decommissionings’ arose again to nearly 30 per cent of the then total by 3300 bce. However, the increase in abandonments thereafter was so sharp that the number of newly redundant sites had reduced to not much more than 10 per cent by 3250 bce, before they had all ceased being used, at least in terms of their initial purposes, by not long after 3200 bce.
The complexity of development and use of some causewayed enclosures should occasion no surprise, given the multiple circuits and multiple indications of recutting of ditch segments indicated at many sites. Windmill Hill near Avebury was mentioned earlier, and the recent close dating programme revealed not only that previous ideas about the sequence of construction of the rings of ditch-segments (and banks) there were correct, but it also established the actual duration of the periods involved. So the inner circuit is now thought most likely to be the first to have been created, its segments having been dug in the period 3670–3645 cal bce. The outer circuit was probably created marginally later (though it could have been started while the ‘final’ segments of the inner were still being cut), sometime between 3670 and 3635 cal bce. And meanwhile the middle circuit appears likely to have been created a little later, having been dug sometime between 3640 and 3615 cal bce. The whole ‘edifice’ was therefore very probably completed in less than seventy-five years, and could easily have been built within an individual (short) lifetime of between twenty and fifty-five years. The complex history of the filling and recutting of the segments was such that activity continued sporadically at the site for a period of 350 years or so, concluding around 3350 bce. However, it is thought that this ‘ending’ really represented a change of use, and some later fourth- millennium deposits, though less common, could still involve substantial amounts of material.
Delimiting space in a linear way
The advent of causewayed enclosures in Kent and Sussex broadly coincided with the emergence of a very different monumental tradition in the lowlands of Scotland. ‘Cursus’ monuments are much less well dated than the enclosures, owing in part to radiocarbon samples that have been dominated by wood charcoal, which may produce less accurate results than bone or antler. They had certainly developed by the thirty-seventh century bce, but may date back into the thirty-eighth. The earliest cursus monuments were elongated rectilinear enclosures composed of numerous free-standing timber uprights or small pits, and therefore defined by a permeable boundary comparable with the causewayed enclosures. They range in size between the sixty-five-metre long example at Douglasmuir in Angus, and the six-hundred-metre Milton of Guthrie cursus, also in Angus. The post-defined examples have been found to have often contained internal divisions, and sometimes very large post-holes or pits were set within the enclosed space. This organization of space invites comparison with the timber halls discussed earlier, and it is arguable that entering a timber cursus would have been redolent of being inside a very large, if roofless, building (Fig. 3.20).
Moreover, like the wooden halls and a variety of other timber structures of the period, it seems that many of these early cursus monuments of northern Britain had been deliberately destroyed by fire. At Holm Farm in Dumfries and Galloway, for example, such a post-defined monument had apparently been burned and rebuilt on a number of different occasions, with the result that some individual post-holes contained the charred stumps of a number of different uprights. Unlike the causewayed enclosures, cursus monuments have few close parallels on the European continent, and it is fairly clear that they developed entirely in the insular context. It is arguable that they evoke the ideas of the house, the household, and the collective history and identity of the domestic group on a massive scale, and it is also evident that they must have been used in an entirely different way from the enclosures. The linear interior of the cursus hints at procession and a sequential ordering of events, while the nested rings of an enclosure are more suited to the containment of gatherings within a privileged space. While, as we have seen, the quantities of pottery, animal bone, and flint found at causewayed enclosures are often very great, post-defined cursus monuments produce only modest numbers of finds. This is in part a result of the contrast between ditches (which may be left open for a period) and post-holes (which are quickly sealed once the post has been inserted) as potential ‘traps’ for objects. But it also arguably reflects another aspect of the contrast between the ways in which these structures were used. The profligate consumption of food and destruction of artefacts were not characteristics of cursus monuments.
It is open to question whether these two different orderings of space could simply be identified with two distinct ‘liturgical’ traditions or systems of belief. These structures are unlikely to have been used for activities that can be narrowly defined in terms of ‘religion’, while metaphysical beliefs may not have been consistent or coherent across time and space in such pre-literate societies. Nonetheless, the two kinds of architecture will have facilitated entirely different kinds of presence, performance, and practice. In general terms, then, it is possible that they represented competing forms of organization of social practices and interaction, and their geographical manifestation and sequence of appearance may shed some light on this. Causewayed enclosures spread rapidly from the south-east as far as Cornwall, Wales, and the English Midlands, and while there are small numbers of examples in Ireland and Cumbria, they are conspicuous by their scarcity in Yorkshire. Meanwhile, pit- and post-defined cursus monuments were concentrated in southern Scotland, so that the two traditions did not initially overlap spatially. However, towards the middle of the fourth millennium bce the character of cursus monuments began to change fundamentally. Pit and post structures ceased being built and at the same time new forms of parallel-sided structures began to appear. These featured continuous ditches surrounding linear spaces, with either squared or curved terminals. This process of change is demonstrated very strikingly at the site of Holywood North near Dumfries, where a post-defined monument was enclosed within a ditched cursus that was more than two hundred metres long (Fig. 3.21).
While ditched cursus often have breaks in the ditches along their sides, they rarely have any means of access at their ends. It follows that it could be argued they are not self-evidently ideal for use as ceremonial routeways. One explanation that has been put forward for this state of affairs is that rather than defining a space down which participants could have processed, ditched cursus monuments were built to enclose an existing pathway and render it inaccessible. The former processional space was, in this perspective, perhaps sanctified, reserved for spirits and ancestors, or even cursed. As such, these may not have been spaces for the living to move within, but instead to venerate or fear. Ditched cursus are much more widely distributed across Britain than the pit and post versions, with major concentrations in the Thames Valley, and in parts of East Anglia, Yorkshire, the Welsh borders and the Midlands. There is a tendency for them to occur on the flat expanses of river valleys. This may simply be because these areas afforded the space necessary for massive linear constructions, but it has also been conjectured that some kind of affinity was being asserted between the cursus and the river. They may, for instance, have been connected in some way with mortuary practices focused on rivers, a suggestion supported by the discovery of rather later Neolithic human remains in a logjam in a former channel of the River Trent at Langford Lowfields in Nottinghamshire. Alternatively, Kenneth Brophy has suggested that cursus may have been understood as equivalents of flowing rivers, embodying ideas of time, linear movement, and mortality. Cursus monuments also often incorporate and link together existing monuments, such as long barrows, ring-ditches, and mortuary enclosures. They are, moreover, sometimes much larger than the earlier post-defined structures: the Greater Stonehenge Cursus is nearly three kilometres long, while the Dorset Cursus on Cranborne Chase extends for a massive ten kilometres.
Despite their sometimes colossal size and the enormous quantities of labour that must have been invested in constructing them, cursus monuments rarely have any internal features, and their ditches contain few if any finds. The Greater Stonehenge Cursus was unusual in this respect, in that the ditch deposits at the western terminal near Fargo Plantation contained numerous clusters of flint-knapping waste. But the character of this working was enigmatic: entire nodules had been flaked until little remained, with few artefacts being produced. It is conceivable that in this context it was the act of knapping itself that was important, rather than the manufacture of any particular object (see Chapter 4 for further discussion). The two cursus monuments in the Stonehenge area, the Greater and Lesser Cursus, are both dated to slightly later than the nearby causewayed enclosure, Robin Hood’s Ball. In the Thames Valley, numerous cursus are again a century or two later than the causewayed enclosures scattered across the gravel terraces. So in some areas, causewayed enclosures and cursus monuments are closely juxtaposed, even if their construction was not precisely contemporary. However, it is entirely possible that particular people would have experienced the activities conducted at both kinds of structure within their lifetimes. Significantly, at Fornham All Saints in Suffolk, a large cursus runs directly across two intersecting causewayed enclosures. A similar relationship was observed between the cursus and causewayed enclosure at Etton, while as we have already seen, at Maiden Castle in Dorset a bank barrow was found to have cut across and involved the levelling of the bank and ditch of the enclosure. Bank barrows and cursus monuments are closely related: indeed, the Cleaven Dyke in Perthshire and the Stanwell Cursus at Heathrow were each effectively a bank barrow contained within a cursus ditch. So given that in several cases it is clear that cursus monuments slighted the earthworks of causewayed enclosures, it seems probable that the practice of creating cursus monuments represented the eclipse of one form of order and way of structuring activity in the landscape, and the emergence of another.
What, therefore, exactly were cursus monuments, if they were not simply meeting places that in some way were equivalent to causewayed enclosures? The near-absence of associated finds would appear to question the parallel with causewayed enclosures anyway, given that the latter seldom feature few finds (although some, like the Mavesyn Ridware and Alrewas enclosures in the Trent valley do seem to have been almost devoid of material). Archaeologists are surely right to have noted that the linear character of cursus enables ‘landscape statements’ to be made, whether restricting or directing passage across the land, linking the locations of former monuments, or pointing to features on hillsides that they were oriented towards. We would argue that we should not necessarily focus so much upon the ‘monumentality’ of these constructions, however, whether in timber or as earthworks, as if that monumentality requires an explanation of ritual or functional importance in its own right. Cursus are perhaps best seen as indicative or mnemonic devices acting dramatically and emphatically within a historicized landscape. Adopting this view, we would accept and extend the argument that Jonathan Last has made that cursus were perhaps developed as a way of ensuring that the idea of a journey or passage became a key metaphor for the societies in question. This metaphor was made manifest in material form, whether the transit was along, or across, a river valley or other significant landscape. As such, the act of constructing the cursus, especially in view of its referencing of past monumental features (frequently long barrows, but also the causewayed enclosures of the then-recent past), was a means of integrating key concepts involving movement and connection (and even the very performance of acts of referencing), in a vivid way. It was the act of creation of a cursus monument that in these terms was important, and not necessarily what subsequently went on within or around it in terms of (for instance) assembly, exchange, or ritual action. That said, once each cursus was in place, it did apparently act as a node within particularly ritually and, one might add, historically charged places in the landscape.
As we have observed already, causewayed enclosures fell out of use during the third quarter of the fourth millennium bce, and it is arguable that in some sense cursus monuments supplanted or superseded them. While there is little evidence for activity within cursus, they seem to have ‘attracted’ subsequent structures and deposits, including ring-ditches, single grave burials, timber circles, henge monuments, and pits containing fine artefacts. This development probably does not indicate one population replacing another, but might just possibly be seen as the eclipse of a largely Continentally inspired set of practices by an alternative that had been generated in Britain. As we have noted, this contrast may not have been as specific as that between two religions or ideological systems, but it may indicate differing ways of organizing or articulating communities. And interestingly, the proliferation of ditched cursus monuments throughout much of the British mainland coincided with a series of other fundamental changes.
Cultural diversities and elaborations
From around 3400 bce onwards, the decoration of ceramic vessels used in Britain began to become appreciably more ornate. The shouldered bowls that had been sparsely ornamented during the Earlier Neolithic now became the core of a new style of pottery, known as Peterborough Ware. Rims became heavier, sometimes amounting to a collar, below which there was often a groove or cavetto zone. Decoration expanded to cover much of the outer surface of the vessel (sometimes expanding into the interior as well), and was usually composed of multiple impressions forming horizontal bands. These were executed using fingertips and fingernails, twisted cord ‘maggots’, and the articular ends of bird bones. In a path-breaking doctoral thesis completed in 1956, Isobel Smith suggested that the development of Peterborough Wares involved a sequence of three styles—Ebbsfleet, Mortlake, and Fengate—in which decoration became progressively more florid and expansive, rims were transformed into collars, and the rounded bases inherited from the Early Neolithic were eventually replaced by flat pedestal bases. There are also differences between the fabrics and finishing techniques of the different substyles that have been recognized. However, as radiocarbon dates began to emerge for deposits associated with pottery in these styles it became clear that there was a good deal of chronological overlap in their manufacture and use. Moreover, as the variability in the different styles has become clearer, and the contexts of deposition have become better understood, it has been recognized as being likely that these different sub-styles were used in slightly different ways. All three styles of pottery often occur deposited in pits, but Ebbsfleet Ware is especially common in causewayed enclosures and round barrows, Mortlake Ware in open sites, rivers, and megalithic tombs, and Fengate Ware also in causewayed enclosures. As Vincent Ard and Timothy Darvill have recently reaffirmed, Peterborough Wares often seem to have been used in locations that were associated with the past: at overgrown and decrepit monuments. This apparent referencing or veneration of places and ancestors many generations old gives the impression that as the British Neolithic had reached maturity, society had become increasingly backward-looking, a tendency that would only be reversed at the turn of the third millennium. The precise significance of the enhancement of decoration on pottery is hard to explain, although ethnographic examples suggest that it may be connected with heightened concern over interpersonal transactions involving the preparation and consumption of food and drink. It is possible, then, that the development of Peterborough Wares has something to tell us about increasingly fraught, or contested, relationships between people of different statuses, ranks, or genders.
If increasingly elaborate ceramics were connected with a need to define boundaries or mark out identities in the casual encounters and activities of everyday life, this harmonizes with a more general growth in the diversity and elaboration of material things in the later part of the fourth millennium bce. A variety of new kinds of portable artefact was created at this time. These included jet belt sliders that were dress fittings suggestive of the emergence of new and more elaborate ways of presenting oneself in public. The skills of grinding and polishing flint were employed to create distinctive new kinds of knife and axe, including the flaked and partially polished axes and adzes of Duggleby and Seamer type. The leaf-shaped flint arrowheads of the Earlier Neolithic were replaced by new, chisel-shaped styles. And maceheads of stone began to be manufactured again, complementing the existing polished stone axes. None of these objects would have been more efficient or helpful in carrying out everyday tasks. All of them were visually arresting, and each would have been very clearly associated with a particular person, worn or carried in social interaction. These items may or may not indicate that social differentiation or ranking had increased by the thirty-fourth century bce, but they certainly suggest that roles and identities were in need of clarification or qualification. British communities may have become more unequal, or there may simply have been a proliferation of statuses and offices in an increasingly diverse and fragmented social world. If the period of the causewayed enclosures was one of more frenetic social interaction, growing competition, and enhanced conflict, its aftermath was perhaps an interlude in which order was reasserted, often by looking back to established traditions and places with ancestral associations.
The descent and diversification of traditions
It is highly salient that many of the new kinds of artefact, suited as they were to display and the expression of personal identity, have been recovered from graves. The practice of burying a single person, crouched in a grave beneath a small round mound, and accompanied by a number of distinctive artefacts, did not represent a sudden and radical break with convention. All the elements of this way of treating the dead already had a long history. Single grave burials had been performed since the start of the Neolithic (as we have seen with the example from Yabsley Street), while round mounds containing multiple bodies had been constructed in a number of regions, notably eastern Yorkshire. Furthermore, the single graves of the later fourth millennium often explicitly drew on existing practices, suggesting continuity with the past. For example, at Barrow Hills, Radley in Oxfordshire, two crouched burials were found in a grave beneath a small long barrow adjacent to the Abingdon causewayed enclosure. But the inhumations dated to the later fourth millennium, and were accompanied by a set of striking artefacts. At Whitegrounds near Burythorpe in Yorkshire, a single male burial with a polished axe and a jet slider was inserted into an existing mound that already contained an Early Neolithic linear mortuary structure (see Chapter 4 for more extensive discussion of these two sites). And at Four Crosses in Powys, the first in a series of burials interred in a round mound over a protracted period was deposited within a two-post structure that harked back to Early Neolithic prototypes.
These Middle Neolithic burials were not distinguished by standardized forms of body treatment or artefact sets, as would be the case with the Beaker graves a millennium later (see Chapter 5). But they do stand out from the long barrow and long cairn mortuary practices that we have already discussed. If the latter had involved the accumulation of a group of ‘founding ancestors’ over a generation or two, from whom an extensive community reckoned their descent, there was now an emphasis on the death and celebration of particular, singular persons. The implication is that new and more exclusive lines of kinship and inheritance were being established in the later fourth millennium, enabling people to identify themselves in relation to a specific, and perhaps relatively recent, ancestor. These tendencies are most spectacularly exhibited at the site of Duggleby Howe on the Wolds of East Yorkshire, where a massive round mound grew up through a series of stages of deposition and construction, associated with a group of burials introduced sequentially over a period of at least half a millennium. The earliest of these burials had been deposited in the thirty-sixth or thirty-fifth century bce, before mound construction had begun, in a deep shaft grave. At the base of the shaft a mature adult male had been laid out, possibly in a wooden coffin, with a pottery bowl in the local Towthorpe style and several flints, including three serrated flakes and two cores. Higher in the same pit was the body of another man, aged about 60, with the skull of another person at his feet, who had apparently died from severe blows to the head. This skull lacked a lower jaw, and had probably been buried in a defleshed state, so the possibility is that it did not represent a burial as such, but a war trophy, an heirloom, or part of an ancestor, retrieved from elsewhere. A little higher in the shaft was the body of a 4-year-old child, and at the top was another man aged about 60, with a macehead worked from red deer antler and a lozenge-shaped arrowhead in front of his chest, and a flint axe with a finely polished blade beside his knees. This body had been buried two or three centuries after the first, demonstrating that the location was one that people had retuned to sporadically in order to bury the dead over a lengthy period even before the mound was raised. The positioning of these bodies in the shaft reveals an intention to create connections between people who lived at different times, so that the burial place embodied the lines of descent passing through the generations. The presence of the separate skull and the antler macehead, which was apparently already a century old when deposited, add to this sense of sequential funerary events as a kind of historical narrative.
The sense of continuity in the use of a specific location was enhanced by the addition of a second, shallower grave beside the shaft, perhaps some time shortly after 3000 bce, and thus more than two centuries after the final burial that the latter contained. The tightly contracted body in the base of this grave was another adult man, with a bone pin, beaver teeth, boar’s tusk blades, and a chisel-shaped arrowhead. These objects significantly prefigure the items that would be found with cremated burials at various sites during the first half of the third millennium. Another body, of a man who may have been as old as 70, was laid out on the land surface between the two graves, had a very fine polished flint knife, and had been placed there perhaps a century later (Fig. 3.22).
This was the last burial before the circular barrow began to be raised over the two graves. The first element of this was a low mound of earth and clay, and four children were buried in this and the layer of chalky grit or fine rubble above it, possibly overlapping with a progressive change in mortuary activity toward the deposition of cremated human remains. The burnt bones of fifty-three people were eventually inserted into the mound, three of them with bone pins similar to that found with the body in the shallower grave. A clay capping and coarse chalk rubble eventually completed the mound. There is no indication of any abrupt cultural change between the two different forms of funerary activity. So although there was a shift towards what might have represented a more spectacular form of funerary activity, with cremation events that might have been witnessed by large numbers of people, the principal emphasis at Duggleby Howe appears to have been on continuity, expressed through the protracted use of a single location. These dead were buried sporadically, with prolonged gaps in the sequence, rather than in an abbreviated episode. This implies the repeated restatement of the importance of a particular genealogical line through time. The recurring presence of quite elderly men amongst the burials at Duggleby suggests that it may have been household or lineage heads that were being afforded this distinctive form of treatment in death, although this may have reflected their position in kin relations as much as their personal authority. Finally, in the twenty-fourth century bce, a massive ditch about 370 metres in diameter was built encircling the mound, and it is likely that the final chalk rubble capping of the barrow represented the up-cast from this feature. It is open to conjecture whether this ditch was intended to enhance the importance of the mound, or to sever it from the surrounding landscape.
Although much later in date, the ditch surrounding Duggleby Howe continued another trend that began in the final centuries of the fourth millennium: the development of circular enclosures, some of which contained earlier features, and which were often associated with the dead. A good example is at Monkton Up Wimborne in north Dorset, where a ring of substantial pits surrounded a much larger central pit, from within which a circular shaft descended to a depth of seven metres. In a secluded location on the opposite edge of the large pit was a grave, which contained the remains of an adult woman and three children, only one of whom appeared to be closely related to her. Strontium isotope values from the tooth enamel of the four people demonstrated that they had moved on at least one occasion to and from an area with a more radiogenic subsoil than the Wessex chalk, probably the Mendip Hills. The burials were radiocarbon-dated to around 3300 bce, and their diet had apparently been exceptionally rich in dairy foods. This is an interesting result, given that, as we have seen, the evidence for the use of domesticated cereals had declined appreciably in southern England by this time. However, it may be that the animal-based diet reflects the mobile way of life of only a segment of society, following herds across the country. A further remarkable discovery from Monkton Up Wimborne was a large block of chalk found at the base of the shaft, which bore pecked linear designs similar to those found on the stones of passage tombs in Ireland and western Britain.
The Monkton Up Wimborne decorated block provides a link with another circular enclosure of the late fourth millennium, at Flagstones House on the edge of Dorchester in Dorset. The Flagstones site had an interrupted ditch similar to a causewayed enclosure of earlier date, although it formed a rough circle about 100 metres in diameter. As we will see later in this book, this feature links the Flagstones enclosure to the early phases of activity at Stonehenge. A series of infant burials were identified in the ditch segments, some of them covered by large sarsen stones. In four places around the enclosure circuit, engraved pictograms were found in the chalk surface of the side of the ditch. The designs included parallel lines, a chequerboard motif, concentric circles, and nested arcs. As at Monkton Up Wimborne, the obvious comparison is with the Irish passage-tomb tradition, which was reaching its zenith at this time. The Monkton and Flagstones ‘art’ provides a tenuous indication that new, longer-distance contacts around the coasts of Britain and Ireland were starting to form by about 3300 bce.
We have seen that the passage tombs of Ireland were probably in existence by the thirty-seventh century bce, and that the largest and most complex examples in the Boyne Valley probably date to the later part of the fourth millennium. Passage-tomb architecture is distinctive amongst megalithic monuments, with chambers lodged inside often circular mounds accessed by long stone passages. This style of construction originated in western France and Iberia, spreading gradually to Ireland, north Wales, northern Scotland, and eventually Scandinavia. Yet while there are distinctive elements shared between regions, there are also local variations in the ways that passage tombs were constructed and used. Thus in Ireland predominantly cremated human remains were deposited in the chambers, sometimes in distinctive stone basins, but in the Orkney Islands the Maes Howe style passage tombs were used for successive inhumation. Passage tombs in Ireland are associated with a style of decorated pottery called Carrowkeel Ware, but this is not present on the Continent, or in Wales or Scotland. One shared element apparently peculiar to passage tombs is the presence of decorative motifs on the stones of passage, of chamber, and sometimes also on those of a surrounding kerb. While the art of the Iberian passage tombs includes painted designs, that of Brittany and Ireland is principally pecked or incised onto the surfaces of the stones, and incision is more common in the Orcadian tombs. Some of the Breton motifs are figurative (of animals, axes, crooks, and bows), but those in Britain and Ireland are predominantly geometric and non-representational, composed of elements including zigzags, nested arcs, concentric circles, parallel lines, lozenges, and triangles. However, this does not mean that they are without meaning. Detailed analysis by Guillaume Robin has demonstrated that particular combinations of motifs are more likely to occur in distinct locations within the Irish passage tombs, including portals and the backs of chambers. While it would be quite misleading to cast the decoration of passage tombs as anything approaching ‘writing’, the carvings nonetheless appear to have drawn attention to the stages of the process of entering the monument, and moving down the passage toward the chamber area.
By comparison to the chambered long cairns of the earlier fourth millennium bce, the encounter with the inner space of passage tombs was a more richly sensory experience. The long passage, which could often not be negotiated in an upright position, was clearly differentiated from the chamber, and separated the mortuary deposit from the outside world. As a result, the chamber was often a dark and secluded space. In some cases the passage had an alignment on a particular celestial event: the interior of the tomb is illuminated by the midwinter sunset and sunrise respectively at Maes Howe in Orkney and Newgrange in Ireland. In other words, particular conditions were being created under which it was appropriate to experience the chamber and the remains that it contained. The simple implication of the effort expended on elaborating the experience of entering these tombs is that they were intended to be occupied repeatedly, over a long period of time. While the funerary deposits in the earlier megalithic tombs had often accumulated in less than half a century, those in passage tombs were usually more extensive, and had been deposited over a longer period. At Knowth, the cremated remains of more than a hundred people had been interred over several centuries, while at the Mound of the Hostages at Tara there were over three hundred burnt and unburnt burials. Similarly, at Quanterness in Orkney, bodies may have been deposited in a fleshed state and subsequently reorganized over a period in excess of 400 years. So although the total number of bodies in these tombs is often large, they may have been introduced relatively infrequently: just a few in any decade. It is worth comparing Duggleby Howe with the passage tombs, for in both cases the deceased were slowly added to a monument that was undoubtedly associated with the unfolding history of a specific social group. This monument did not celebrate a group of collective ancestors in the relatively distant past, but embodied a specific line of descent, traced from past to present. With the emergence of developed passage tombs and single grave burials, the relationship between the living and the dead had begun to shift, apparently decisively.
