The Thames Valley in the late fifth and early fourth millennium cal bc: the appearance of domestication and the evidence for change
The Thames Valley in the late fifth and early fourth millennium cal bc: the appearance of domestication and the evidence for change
Abstract and Keywords
This chapter reviews the evidence for the late fifth and early fourth millennia cal
IN THIS PAPER, we review the evidence for the late fifth and early fourth millennia cal BC in the Thames Valley, a time of profound change which is traditionally studied by specialists from two very different schools of research. Our task is made more challenging because many sites of this period tend to be small and appear to have been short-lived. As a result, there are few strati-graphic sequences present which would lend themselves to more precise statistical analyses of radiocarbon measurements. In addition, sites with fifth millennium cal BC assemblages are not common and are not easy to interpret.
Nevertheless, our interpretation of this period in the region has altered dramatically over the last 15 years. This has come about largely as the result of development-led archaeology and the resulting shift in focus of investigations from areas with well preserved monuments such as Avebury and the Cotswolds to the now well populated river terraces and floodplain (Allen et al. 1997; Cotton & Field 2004; Holgate 1988). Extensive area excavations have taken place on the gravel terraces, for example around Heathrow, Eton and Maidenhead in the Middle Thames (Allen et al. 2004; Ford et al. 2003; Lewis & Welsh 2004) and Yarnton and Shorncote in the Upper Thames (Barclay et al. 1995; Hey in prep.; Laws 2004). These have been supplemented by hundreds of small- and medium-sized investigations and evaluations. While this has resulted in an explosion of new information about site types and distributions, and a greater understanding of the wide range of activities in which people were engaged in the past, the sites have largely been of fourth millennium cal BC date with relatively few sites of the fifth millennium cal BC.
Our understanding has also been greatly enhanced by the development of palaeoenvironmental techniques and better radiocarbon dating using a more (p.400) critical, integrated approach based on Bayesian modelling (e.g. Bayliss et al. 2007; see also Whittle, this volume). The results of this work are likely to have a profound impact on our understanding of the period under discussion here and will allow for a more precise chronology, perhaps based on timescales as exact as human generations.
Time and Sequence
There are few late fifth millennium cal BC radiocarbon dates from the Thames Valley, and most of these are single dates or assays on poor sample material (Barton & Roberts 2004, 345–7; Schulting 2000; Williams 1989), for example the date on charcoal from a hearth at Wawcott Site I in the Kennet Valley (BM-449: 5260±130 BP, 4350–3750 cal BC; Froom 1971–2, 27). There are many new dates from the Lower Thames and Thames Estuary on peat and organic-rich deposits (Bates & Whittaker 2004, table 6.2 and appendix I) and some dates are beginning to be acquired during work in the London area (Sidell & Wilkinson 2004), but few of these can be directly associated with human activity. It is possible to say that, at present, there are no well dated hunter-gatherer sites of the late fifth millennium cal BC, and no evidence for domesticates, pottery or monuments before 4000, possibly even 3900 cal BC (Barclay 2006; Whittle et al. 2007).
Charred cereal remains of early fourth millennium cal BC date have been recorded beneath the Hazleton North long cairn, from the Area 6 midden at Eton, a pit deposit at Yarnton, from Woolwich, and from the longhouse at White Horse Stone in the Medway Valley, Kent (Allen et al. 2004, 91; Bates & Whittaker 2004, 67; Bayliss & Hey in prep.; Hayden forthcoming; Meadows et al. 2007). Some cereal pollen was also recorded from a basal clay deposit at Daisy Banks, just outside the Abingdon causewayed enclosure. A radiocarbon date of 4350–3750 cal BC (OxA-4559: 5240±110 BP; Parker 1999, 260) was obtained on charcoal and seeds from the bottom of this sequence. Immediately above this layer, there was evidence for major cereal cultivation which may be contemporary with the use of the causewayed enclosure (Adrian Parker, pers. comm.).
Occupation sites and middens with pottery, domesticates, polished stone tools and buildings appear between 4000–3800 cal BC; monuments are also constructed at this date but only in a few restricted areas, such as the Cotswolds and the Medway Valley. The greatest phase of monument building (long barrows, causewayed enclosures, mortuary enclosures and cursus monuments) appears to belong to the mid fourth millennium cal BC, c. 3650–3350 cal BC (Barclay 2006; Whittle et al. 2007).
The Thames Valley is one of the great river systems of Britain. It extends over 150 km across southern England, and its mouth opens out on to the North Sea and the lowlands of Western Europe, while its headwaters reach into the Cotswold Hills, the Chalk Downs and the Avebury area. It contains a number of diverse and well known areas, including the Thames Estuary and its foreshore, Avebury and the Kennet valley, the Upper and Middle Thames gravels, the Berkshire Downs and the Cotswolds (Fig. 1; Holgate 1988).
In the fifth millennium and at the start of the fourth millennium cal BC, a number of significant palaeoenvironmental changes occurred in this area. A marine transgression around the Thames Estuary was followed by a drop in sea level from around 4500 cal BC, allowing the formation of peat and eyots within the river channel (Bates & Whittaker 2004; Siddell & Wilkinson 2004). A fluctuating fluvial regime in the Lower Thames may have made the flood-plain less attractive for permanent settlement (Siddell & Wilkinson 2004, 40–1).
The elm decline is also evidenced at a number of sites at around 4000 cal BC (Robinson 1999, 269–70; 2000a, 30–2), although it is not always apparent in sequences spanning this period, for example at Runnymede or in parts of central London (Robinson 2000a; Siddell & Wilkinson 2004, 41–3). Sometimes this event seems to be coincident with woodland clearance, as at Daisy Banks, Abingdon (Parker 1995), but it is also apparent in places where clearance did not take place until the late Neolithic/early Bronze Age, for example at Sidlings Copse, north-east of Oxford (Day 1991, 465–7). The link between the elm decline and human activity, in particular clearance for agriculture, remains unproved.
Evidence from the length of the Thames Valley shows that, by the fifth millennium cal BC, the area was covered by mixed deciduous woodland, with alder growing in the valley bottoms and lime, oak, hazel, ash and elm on better-drained soils of the gravel terraces and higher slopes (Day 1991; Robinson 1992, 49–50). Forest cover had become more dense over the course of the previous 3000 years. This picture emerges from a range of pollen and macrobotanical studies stretching from the Cotswolds to the Thames Estuary (Bates & Whittaker 2004; J. Evans et al. 2006; Robinson 1993, 12–14; 2000a; Siddell & Wilkinson 2004).
At Runnymede, environmental evidence indicates some clearings in the fifth-millennium forest in the form of open woodland within more dense alder carr (Needham 2000, 193–5), and a similar pattern is suggested at Thatcham (Healy et al. 1992, 70–2). The extent to which hunter-gatherer populations either took advantage of natural clearings where a wider range of plant foods would be available, or created and maintained these deliberately has been widely discussed (e.g. Tipping 2004), but the close association between such environments and hunter-gatherer sites seems inescapable (Bell & Walker 1992, 156–8; Zvelebil 1994).
Human interference in this forested landscape is clearly demonstrable from the beginning of the fourth millennium cal BC. Where the evidence exists, however, sites of this date are almost always in small woodland clearings, for example Ascott-under-Wychwood and Hazleton North long cairns (Benson & Whittle 2006; Saville 1990). Woodland recolonisation subsequently occurred on both sites. An early Neolithic long enclosure of arguably slightly later date at Yarnton appears to have had a similar history of clearance in advance of construction followed by woodland regeneration. The general vegetation model for this area in the Upper Thames is one of shifting clearings in woodland against the background of the gradual opening out of the canopy (Hey & Robinson in prep.). A little further downstream, the Drayton Cursus was constructed in a cleared landscape where earlier monuments were present, but there were several phases of regeneration and clearance in the middle and late Neolithic (Barclay et al. 2003).
At Eton in the Middle Thames, there is no evidence for widespread clearance in the early fourth millennium cal BC, despite the presence of middens and other contemporary activity nearby (Tim Allen, pers. comm.). A similar picture emerges at Runnymede, where pollen dating to the early fourth millennium cal BC indicates deliberate interference in tree canopy, although plant macrofossils continue to show dense alder woodland (Needham 2000, 193–5; Robinson 2000a, 31–2). In the London area, land clearance is associated with cereal cultivation on some sites in the early fourth millennium cal BC, although adjacent areas appear to have remained wooded (Siddell & Wilkinson 2004, 42).
Woodland clearings evidently were important to populations of the fifth millennium cal BC in the Thames Valley, and may have been augmented and maintained by local groups to maximise resource aggregation. This would have provided opportunities for larger-scale gatherings of what may, on a day-to-day basis, have been small groups of people. The extent of clearance increased visibly in the early fourth millennium cal BC, and the scale and (p.404) character of this activity show that this was deliberately undertaken and maintained. Nevertheless, few clearings were long-lasting; rather they shifted across the landscape. The overall impression of the landscape of the mid fourth millennium cal BC is of a mosaic of woodland and clearings (Robinson 2000a, 33), but with few permanently open areas. Where more extensive grassland existed this lay next to monuments which were the scenes of large-scale gatherings, such as causewayed enclosures and cursuses (Barclay & Hey 1999).
Settlement in woodland
Within this forested environment, evidence of fifth and fourth millennium cal BC activity is surprisingly widespread in the Thames Valley and its catchment (Holgate 1988). Sites of this date, however, are often found as artefact scatters and are difficult to interpret as they are without a precise context and rarely associated with environmental evidence.
In 1988, Robin Holgate suggested that activity dating to the end of the fifth millennium cal BC tends to be found on higher ground away from the valley floors where the earlier sites were often situated. More recent work in river valleys has shown that there are many more sites in lower topographies than he predicted (see below), but does indicate that later hunter-gatherer sites in these environments are smaller in size, suggesting that people moved in smaller groups or stayed in one place for shorter lengths of time. Increasingly dense forest cover, as witnessed for example at Eton (see below), may have reduced the number of locales with large concentrations of plants and animals and may have made smaller groups more effective at resource maximisation, as suggested for other valley systems in Europe (Whittle 1996, 153).
Many sites on higher ground are also small, and Holgate has argued that the restricted range of implements present and the high proportion of microliths (with few or no tranchet axes or axe-sharpening flakes) indicate that these represent seasonally occupied hunting camps (Holgate 1988, 74–6). Based on both ethnographic and archaeological studies, Spikins (2000) has been critical of very simplistic models of Mesolithic land use. She points out the wide variety of potential uses of sites, and the adaptability and flexibility of hunter-gatherer groups in their use of the landscape. Additionally, use-wear analysis has shown that microliths were used for various tasks, not just hunting (Grace 1992, 60–2). Recent workinthe region has demonstrated that, away from valley bottoms, sites of varying sizes existed on which a range of activities seem to have taken place, and the site at Windmill Hill near Nettlebed is a good example of this (Boismier & Mepham 1995). It may be better to think of Mesolithic groups as highly knowledgeable exploiters of their environment, (p.405) moving across the landscape according to tradition, taking advantage of seasonal and other natural resources but also of more occasional fluctuations in the availability of different foods. Occasional abundance may have provided opportunities for gatherings of larger numbers of people. In the light of these new ideas, a review of fifth millennium settlement is overdue.
Settlement of the early fourth millennium cal BC is also widespread and, although the pattern is more dense than in the fifth millennium, sites are similarly small and appear to have been short-lived. There is remarkably little evidence for occupation of any great duration and certainly little to suggest that individual settlement events lasted for more than a few months except in rare instances. On the other hand, evidence for the reuse of the same site on an episodic basis is rather common. Small discrete groups of features at Yarnton provide good examples of repeated, but short-lived visits to one place (Hey 1997, 106–8).
Fifth and fourth millennium cal BC sites also share a similar distribution. Thrupp, Corporation Farm and Gravelly Guy are examples of sites with evidence of fourth millennium use where there had previously been hunter-gatherer encampments at the forest margin overlooking the floodplain (Holgate 1988, 87; 2004a). At Hazleton and Rollright, fourth millennium activity was preceded by smaller-scale episodes of use, perhaps for hunting (Lambrick 1988, 111–12; Saville 1990, 240), and a similar pattern is present in the Upper Kennet Valley (Pollard 2005; Whittle 1990). The juxtaposition of Mesolithic and early Neolithic sites has also been noted by Field for the Thames further downstream and on the Greensand of the Weald to the south (Field 2004, 156).
The river and the riverbank
At a time of dense woodland, the Thames and its tributaries would have been major routeways, as well as environments which provided a rich and varied plant and animal food resource (Clarke 1976, 464–5). The importance of these waterways is symbolised by finds of these periods which have been recovered from the river, for example Mesolithic picks and axes downstream of Goring and similar finds from London, including two bone harpoons (T. G. Allen 1995, 117–18; Field 1989; Haughey 2000, 225–8). An important collection of polished stone axes and some early Neolithic pottery have also been found for which a votive interpretation is probable (Bradley 1990, 66–7; Holgate 1988, 283–4 and 311–35). There is slight evidence that the practice of (p.406) depositing human remains in rivers began during the early Neolithic (Allen et al. 2004, 97; Bradley & Gordon 1988, 508).
It is noteworthy that many fourth millennium cal BC sites in the area lie close to the river. This is particularly marked along the Middle Thames at Eton, Cippenham, Bray Weir Bank Stud Farm, Bray Marina, Cannon Hill and the Maidenhead Flood Alleviation Scheme, where sites cluster in an area of 6 by 4.5 km near to the river (Fig. 2; Allen et al. 2004; Barnes & Cleal 1995; Bradley et al. 1981; Ford & Taylor 2004; Holgate 1988, 278). In contrast, little has been found in survey on the brickearths to the west of Slough (Ford 1987; Ford & Taylor 2004, 99).
At Yarnton in the Upper Thames, early fourth millennium cal BC sites were situated less than 0.5 km from the river on gravel islands in the flood-plain, and early sites on the Cotswolds are often at the heads of tributary valleys, for example Ascott-under-Wychwood and Rollright. The continuing importance of these arteries can be seen in the positioning near river confluences of causewayed enclosures and cursuses in the middle and late fourth millennium cal BC respectively (Barclay & Hey 1999, 68–70).
(p.407) Fifth millennium cal BC sites are often found in similar locations. In the Kennet, major sites have been found on or near the valley floor, including Wawcott and Thatcham, where the range of implements suggests either permanent occupation or seasonal use with a range of tasks being performed (Holgate 1988, 98). At Eton, by 5000 cal BC, lakes and reed fen on the flood-plain had silted up and a series of channels flowed through the area on the banks of which levées had formed. Alder carr developed over the back-swamps (Allen et al. 2004). Fifth millennium flint scatters have been discovered which, although not dense, were widespread and were mainly found on the levées close to the channel, but sometimes stretching back on to the flood-plain perhaps indicating trails leading through the forest (Tim Allen, pers. comm.). In the early fourth millennium cal BC, activity was more widespread than previously but still mainly followed the channel edge, and exhibited a similar pattern of land use (Allen et al. 2004, 85).
Although few sites have been investigated over such a scale as Eton, continuity in the location of settlement in riverside locations is apparent elsewhere, for example along the Maidenhead Flood Alleviation Scheme (Allen et al. forthcoming). At Runnymede, traces of late Mesolithic and early Neolithic activity were present on a silt island between two river channels, on an area later covered by middle Neolithic features and in situ refuse (Needham 2000). Although the earlier material remains were not dense, the evidence suggests repeated use of the site throughout this period of time (Needham 2000, 71 and 240). The proximity of early Neolithic to Mesolithic sites has also been noted further down the Thames in the Estuary (Bates & Whittaker 2004, 59; Field 2004, 156).
At Yarnton, the picture is more ambiguous. Later Mesolithic flint was recovered during fieldwalking over the central gravel island on which early fourth millennium cal BC features were present from a range of activities. It was found in very small quantities, but its correlation with areas of early Neolithic activity is marked.
Zvelebil drew attention to the sophisticated use of wild plant resources by hunter-gatherer groups, ranging from opportunistic and incidental plant use to tending and managing wild resources, including burning the woodland (Zvelebil 1994, 37–40). Plant acquisition could have been one of the factors which influenced settlement strategies. There is, however, very limited evidence of plant use in the Thames Valley by hunter-gatherer communities in the fifth millennium cal BC, except for the ubiquitous presence of hazelnut shells on sites with suitable preservation such as Thatcham (Scaife 1992).
(p.408) The Neolithic is often taken to be synonymous with cereal cultivation and, as discussed above, cereals appear in the earliest dated fourth millennium cal BC contexts in this region. Intriguingly, most of the evidence for Neolithic cereals comes from contexts of the first half of the fourth millennium cal BC; later Neolithic cereal remains are sparse. At Yarnton, cereals formed around a third of early fourth millennium charred plant foods, whereas in middle and late Neolithic contexts approximately 1% of charred foods retrieved were cereals, the vast majority being gathered wild foods, as is typical of pit deposits in the area (Moffett et al. 1989; see also Bogaard & Jones, this volume). A similar pattern is mirrored at Eton and other sites in the region (Pelling forthcoming; Robinson 2000b, 207).
The wide range of edible plants in river valleys and estuaries would have attracted wild animals (Clarke 1976, 464–5). Holgate thought that a shift in the fifth millennium away from marsh and riverside locations in favour of deciduous forest on upper slopes suggested a decrease in fishing and fowling and increased dependence on ungulates. Certainly at Stratford’s Yard, Chesham, all animal bone recovered was of ungulates (wild cattle, red deer, roe deer and wild boar; Grigson 1989) and a similar assemblage was present at Wawcott IV and XXIII (Holgate 1988, 98). However, fishing and fowling are notoriously difficult to detect in the archaeological record, and the absence of fifth millennium cal BC sites in low-lying areas is, as we have seen, not borne out by recent evidence.
Domesticated animals—cattle, sheep and pigs—are present in all kinds of deposits of the early fourth millennium cal BC, along with deer and other hunted animals (Serjeantson 2003). Lipids on pots from Ascott-under-Wychwood, Yarnton, Eton and the Maidenhead Flood Alleviation Scheme show the presence of animal fats from both ruminants and pigs, and also the residues of dairy produce (Copley et al. 2003).
At Runnymede, although the alder woodland canopy was largely intact in the early fourth millennium, there was a sharp rise in the number of dung beetles showing concentrations of large domesticated herbivores, probably cattle and pig, grazing in the forest (Robinson 2000a). The evidence for herding as an important activity increases through time, and woodland clearance may have been as much, or more, to do with creating browse as plots for cereal cultivation. If early communities were small in size, turning the wooded environments into open grassland may have been a gradual process.
In this woodland environment, tree-throw holes are frequently present on sites, and it is not uncommon to find pre-fourth millennium material within them. Tree-throw holes at Gatehampton Farm, Goring, yielded almost entirely Mesolithic flintwork and, on some parts of the site, exclusively later Mesolithic material, even though activity from the early fourth millennium cal BC onwards was present (Brown 1995, 80–1). Substantial assemblages from features at Charlwood in Surrey were probably also mainly from tree-throw holes (Ellaby 2004, 15–16 and see sections on fig. 2.3), as were those from Farnham, Surrey (Clark & Rankine 1939). Additionally, some pre-fourth millennium tree-throw holes contain material that appears to have been deliberately placed there, for example those at Eton (Allen et al. 2004, 91).
The majority of sites at Eton and on the Maidenhead Flood Alleviation Scheme had evidence for the discard of early fourth millennium material in tree-throw holes (Allen et al. 2004, fig. 9.2 and 91; Allen et al. forthcoming). At Eton, middened material was found that is indistinguishable in character to that found in hollows on the adjacent ground surface. Excavation of the Drayton Cursus revealed a number of tree-throw holes that were filled with burnt and redeposited material which appears to derive from the household, but with some placed deposits representing more special activity (Barclay et al. 2003, 60–7). Deliberate use of tree-throw holes has been proposed for other parts of southern England, as a means by which people registered occupation events within a natural forest environment and in the context of shifting settlement (C. Evans et al. 1999). At Eton and Maidenhead it is possible to interpret a number of these deposits in the same way, perhaps marking the temporary abandonment of an occupation site.
Thus, deposition within tree-throw holes appears to be current before the fourth millennium cal BC and continues into the later Neolithic period. However, as pit deposits become more common through the fourth millennium cal BC, the use of tree-throw holes for the discard of material starts to diminish.
One activity which persisted in the Thames valley throughout the period under study was the creation of middens. Middens are a well-known phenomenon of the Mesolithic period and are found in Thames Valley contexts, mainly in low-lying elevations where there has been better preservation, for example Wawcott and elsewhere in the Kennet Valley (Holgate 1988, 4).
(p.410) Early fourth millennium cal BC middens are also present and the spreads discovered at Eton, in an area where no contemporary structures have been found, are of particular interest. They had their origin in the early fourth millennium cal BC and seem to have had a main period of use of around 200–300 years (c. 3800–3500 cal BC), although they may have lasted for as long as 600 years. The generally highly fragmented pottery, struck flint which had been much reused and soil micromorphology indicating considerable trampling over the midden areas, suggest frequent visits to these sites which may have been areas of provisional discard (Allen et al. 2004, 85–91). This does not seem to be casually deposited material. Items had been collected and brought to these places, perhaps from temporary middens, though whether this was the residue of special events or everyday life is uncertain.
There were three middens at Eton, one up to 80 m long, lying within 3 km and all close to the river, in an area with much other contemporary activity (see above). The middens seem to represent episodic but repeated use, perhaps 60 or so individual episodes of deposition over up to 600 years. A similar pattern of periodic use has been suggested for the Mesolithic and early fourth millennium activity at Runnymede (Needham 2000, 240).
The middens encountered beneath the long cairns of Ascott-under-Wychwood and Hazleton North, in the entirely different geographical location of the Cotswold headwaters of the Thames, were situated in places that had been visited by hunter-gatherer groups and were later used for funerary/ceremonial activity.
At Hazleton, the abraded and fragmentary character of the finds and the occasional refits scattered over the area of the midden suggested redeposition of rubbish accumulated elsewhere (Saville 1990, 240–1). Material included a wide range of pottery vessels indicating drinking, cooking and storage, cultivated cereals, many hazelnut shells and evidence for the slaughter and consumption of sheep, cattle and pigs nearby. Some human bone was also present. The site then appeared to have been ploughed. It was suggested that this material represented domestic activity overlain after a period of time by the construction of the cairn; whether deliberately or because the area was still relatively clear of trees was uncertain (Saville 1990, 254).
Recent post-excavation work on the Ascott-under-Wychwood long barrow suggests that the creation of the midden there was very similar. After a short interval, the midden seems to have been linked in deliberately to the construction of the monument and to the activities that accompanied the foundation of the monument (Benson & Whittle 2006). Analysis of the finds suggests that the midden represents the remains of gatherings of people from a wide area, at which the presentation and sharing of food was an important act. The physical traces of the midden would have been a (p.411) mnemonic of these events. Links between deposits of curated material and later monuments have also been noted in the Avebury area (Pollard 2005, 109–11; Whittle 1990, 107).
Pollard has suggested that such sites were projects of accumulation. They would have acted as visible markers of past occupation events and, as they grew in size and age, they would have gained traditional significance (Pollard 1999; 2005; Saville 1990, 254–5). Additionally, they would have been highly fertile places which would have encouraged good plant growth, including plants grown from seeds that had been selected and then discarded on the midden. This may have enhanced their symbolic significance, perhaps conveying notions of fertility and regeneration, as well as making them more distinctive features in the landscape (Bell & Walker 1992, 112; Whittle 1990, 107). Apparent cultivation over the midden at Hazleton North shows that this characteristic was recognised by early populations.
The creation of middens, their augmentation through time, their representation of the past and connotations of abundance and fecundity seem to have been features of both fifth and early fourth millennium cal BC populations in the Thames Valley, although this activity diminishes through the fourth millennium. They demonstrate that people made repeated visits to particular locales, probably not on an annual cycle but as part of a traditional routine, and that these patterns of activity were as important to early farmers as to earlier hunter-gatherer groups.
Another indicator of the persistent use of particular places in the landscape through this period of time is the presence of pits. Pits have been found on a number of fifth millennium and earlier sites (Mithen 1999, 43). In the Thames Valley, four possible small pits or postholes were found at Stratford’s Yard, Chesham, associated with layers containing struck flint including microliths and microburins and animal bones (Stainton 1989). A possible marker post, 0.85 m deep, was found at Runnymede which was reminiscent of the eighth millennium cal BC post pits found in the Stonehenge car park (M. Allen 1995, 43–7; Needham 2000, 193), and pits or tree-throw holes were also found at Charlwood, Surrey (Ellaby 2004).
Nevertheless, pit digging becomes a much more common practice in the fourth millennium cal BC and, in contrast to the formation of middens and deposition in tree-throw holes, it becomes more frequent through time. The extent of pit groups in the area has become more apparent over the last 20 years, as the number and scale of excavations have increased. In 1988, Holgate was able to list 60 Neolithic pits in Oxfordshire, but since that time hundreds have been excavated: over 200 at Yarnton alone.
(p.412) Pits sometimes cover extensive areas as at Yarnton, Lake End Road on the Maidenhead Flood Alleviation Scheme, Drayton and Gravelly Guy, and their date ranges indicate repeated use of individual areas over long periods of time (Allen et al. forthcoming; Barclay et al. 2003; Hey in prep.; Lambrick & Allen 2004). Such sites tend to be in river valleys, often quite close to the Thames or its major tributaries. Elsewhere, pits are found as tightly defined clusters, for example at Benson in South Oxfordshire (Pine & Ford 2003) and South Stoke (Timby et al. 2005), perhaps indicating more intensive but less long-lived use of single sites. Very often, however, they are recovered as isolated features or in pairs, perhaps indicating a single visit, and the geographical distribution of these is widespread (e.g. Booth & Simmonds forthcoming; Brady & Lamdin-Whymark forthcoming). It is rare to find the dense clusters of pits that are a feature of early fourth millennium sites of eastern England, such as those investigated at Spong Hill (Healy 1988).
Excavation at Yarnton indicates the huge variation in the character of the deposits found within these features; they all have their own individual signature. Many pits contained deposits resembling middened material, with characteristically burnt or blackened soils, highly fragmented and abraded pottery with few refits, flint (usually in much fresher condition) and food remains. Other pits have more highly structured contents, including deposits with very large pieces of pottery from decorated vessels in a wide range of shapes possibly representing feasting sets or pits with fine flint objects (Hey in prep.).
In general terms, the pattern of deposition within pits is not one of mundane rubbish disposal, but neither are these features suggestive of large-scale ceremonial events. Most are probably the result of small-scale household rituals commemorated by deposition in the ground, the lasting symbol of individual and personal events (Hey et al. 2003). They could represent the end of individual phases of settlement on particular sites, as suggested for deposits in tree-throw holes in Eastern England (C. Evans et al. 1999). The very careful way in which small amounts of material were selected for use in many pits raises the possibility that these were token deposits, representations of pots, for example, rather than the whole pot itself. Cremated human bone is also found in small quantity, sometimes only around 10 g, and this too could be symbolic.
In common with much of the British Isles, no fifth millennium cal BC human remains are known from this area (Schulting 2000). Formal deposits of human bone have been found in early fourth millennium contexts, however, although these are only rarely associated with monuments. One of the most (p.413) striking recent discoveries has been that of a single burial found on the sandy banks of the river Thames at Blackwall, Greater London, within an oak-lined grave from the charcoal of which an early date has been obtained, though this gives only a terminus post quem for the context (KIA-20157: 5252 ± 28 BP; 4230–3970 cal BC; GLAAS 2004). A large part of a Carinated Bowl had been placed over the head and another Carinated Bowl was recovered from a nearby pit.
Human bone was included as part of a foundation deposit in the Yarnton longhouse (Hey in prep.). Burial within the Cotswold Severn long cairns, and possibly the Medway megalithic burial chambers, also began in the early fourth millennium, but most funerary monuments belong to a slightly later date (Whittle et al. 2007). The early dates suggested for portal dolmens in this area have not yet been confirmed by scientific dating (Darvill 2004, 50).
About the House
Another new feature of fourth millennium cal BC activity in the area is the presence of houses. There are rare examples of dwellings in Britain predating the fourth millennium, including one that has been claimed at Broom Hill, Hampshire (O’Malley & Jacobi 1978), but no such structures have been recognised in the Thames Valley.
One of the exciting developments of the last ten years has been the discovery of early Neolithic buildings in the Thames and Medway valleys. Structures of this date were already known from Gorhambury, Hertfordshire, and Sale’s Lot, Gloucestershire (Holgate 1988, 111–13; Neal et al. 1990; O’Neal 1966), but the new discoveries still only bring the number of substantial houses found in lowland England to around eight. Radiocarbon dates for the houses at Yarnton and White Horse Stone indicate that both were built and used between 3900–3700 cal BC and the same appears to be true of Sale’s Lot.
Both the Yarnton and White Horse Stone structures were set in small woodland clearings in areas with only slight evidence for pre-fourth millennium activity (Hayden forthcoming; Hey in prep.). The White Horse Stone and Pilgrim’s Way buildings lay in the same dry valley and may have been intervisible. The Yarnton house, on the other hand, stood in relative isolation as shown by widespread stripping of the surrounding gravel terrace (Hey & Bell 1997, fig. 8).
The sizes, shapes and designs of these buildings vary greatly (Fig. 3). The longhouse at Yarnton, was made up of a basic rectangle, 21 by 11 m, divided into two modules with some substantial post-pits which presumably supported the roof. It also had outer lines of smaller posts which may suggest a trapezoidal outer shape and a maximum width of 15 m (Hey in prep.; Hey & Bell 1997, fig. 9). The aisled building at White Horse Stone in the Medway Valley was more clearly rectangular in shape and slightly narrower, being some 8 m wide and 20 m long; parts of its outer long walls were defined by slots (Hayden forthcoming). A smaller structure found nearby at Pilgrim’s Way was less well preserved with only postholes surviving, but its similarity to the deeper posthole arrangements of the White Horse Stone house is striking. Several interpretations have been attempted of the possible building beneath the Sale’s Lot long cairn (Darvill 1996, fig. 6.5; Holgate 1988, fig. 6.19), but it is possible to see the gullies, hearth and postholes as belonging to a building like those at Yarnton or White Horse Stone (Barclay 2000). All these houses belong to a common architectural theme of large rectangular roofed timber structures (Darvill & Thomas 1996).
(p.415) The structure at Gorhambury, Hertfordshire was smaller, although it too was rectilinear (Neal et al. 1990, 8–9). It was defined by gullies and was constructed in separate square modules, having parallels with the house discovered at Fengate (Pryor 1974) and some Irish examples (e.g. Logue 2003). The range of early Neolithic buildings types is expanded by a circular structure at Yarnton which has recently been radiocarbon dated to the second quarter of the fourth millennium cal BC (c. 3600 cal BC; Bayliss & Hey in prep.). Other structures at Windmill Hill, Ascott-under-Wychwood and Hazleton North (Benson & Whittle 2006; Saville 1990; Whittle et al. 1999) are less recognisable as buildings and can be interpreted as façades, fences and screens.
Whether all these buildings had a common function and were used throughout the year is uncertain, as their large size should not necessarily be taken to indicate permanent settlement (Whittle 2003, 40–4; see also Bradley, this volume). Consideration of the finds from them is not obviously helpful. With the exception of deliberate packing material in one post pit, the Yarnton building was virtually clean of material, even though the soil from all the post-holes was sieved. At White Horse Stone, the postholes were found to contain fragmentary material: charred plant remains, charcoal flecks, flint chips and small fragments of animal bone and pottery. Sale’s Lot and Gorhambury yielded similar finds, and all these latter structures appear to be buildings in which small quantities of material accumulated and/or were discarded and were swept into postholes. The general absence of material remains around structures may be explained by careful collection and deposition elsewhere, for example in middens, although these have not yet been found close to houses. Traces of middens at Yarnton were around 450 m from the house, and no midden deposits were found near the White Horse Stone and Pilgrim’s Way structures despite extensive stripping of the surrounding areas.
It has been argued that these large structures were not domestic dwellings, but were cult houses or halls for feasting (Thomas 1996). Their size alone, (p.416) perhaps, hints that they were communal buildings able to accommodate more than a single family group. Nevertheless, the structures excavated so far in this area have yielded no obvious feasting debris. Human remains were found associated with the Yarnton structure, however, and a link with mortuary and/or cult activity is possible, although human bone might not have been out of place in any dwelling of this period. Given the range of evidence, it may be more appropriate to think of these buildings as serving more than one purpose and being the locales of formal activities within the domestic sphere (Bradley 2003; this volume).
Throughout the period under study, there are strong strands of continuity. The utilisation of tree-throw holes, the small-scale digging of pits, the creation and abandonment of occupation spreads, and the accumulation of occupation material into middens are common to both periods. Holgate (2004b) has noted that the method of working flint is ‘virtually indistinguishable’ throughout except that, at some point, microliths were no longer produced and new tools such as polished axes and leaf-shaped arrowheads appeared. There is a pattern of the recurrent use of particular places in the landscape throughout. The evidence does not seem to be of continuous activity, not even annual events, but seems to represent episodic, repeated visits to sites which had been cleared in the past, may have been marked in some physical way by middens or posts, and which had special significance to communities who lived in the area. People throughout this period shared a common landscape experience of a largely wooded environment in which the river was a dominant feature. They inhabited woodland clearings which were often close to the river or its tributaries.
However, in the fourth millennium cal BC, communities began to alter their landscape through increasingly substantial building projects: first houses and then monuments. There was more visible treatment of the dead and deposition of human remains. Clearings became more extensive, perhaps largely for pasture, and small cultivation plots were created.
Cereals and domesticated animals, new flint tools and Carinated Bowl are found on all sites from the beginning of the fourth millennium cal BC. Such material is encountered on what can be regarded as traditional sites, for example in midden deposits, and in new constructions such as houses and monuments.
It is tempting to try to rationalise this evidence into explanations of either indigenous populations adopting a new way of life, using the evidence of (p.417) continuity (which is strong); or incomers, pioneer farmers, bringing their own material culture and different social practices, as witnessed by the new elements in the archaeological record (which are striking). But perhaps we should not be thinking in terms of either/or but, rather, both. Is it not possible that sites such as White Horse Stone and Yarnton represent the settlements of new people moving into a landscape that is lightly populated by small hunter-gatherer groups? In the case of Yarnton, they may have used land that had been cleared previously, but which was not frequently used, whereas at White Horse Stone new clearings seem to have been created in virgin forest. Existing communities, for example those around Eton, may have continued to inhabit the area following traditional land-use practices. In this context, newcomers may have been few but have had a profound impact on local groups, both physically and socially, leading them to reconsider their identity and social values and, ultimately, to change at least the ways in which traditional values were expressed in the material world.
Note. We would like to acknowledge the considerable help of Tim Allen and Chris Hayden of Oxford Archaeology, Mark Robinson of Oxford University, Adrian Parker of Oxford Brookes University and Helen Glass of Rail Link Engineering, in providing information and discussing ideas. We are also grateful to Alasdair Whittle and Vicki Cummings for their encouragement and patience during the production of this paper.
ALLEN, M. 1995. Before Stonehenge. In R. M. J. Cleal, K. E. Walker & R. Montague, Stonehenge in its landscape: twentieth-century excavations, 41–62. London: English Heritage.
ALLEN, T. G. 1995. Lithics and landscape: archaeological discoveries on the Thames Water pipeline at Gatehampton Farm, Goring, Oxfordshire 1985–92. Oxford: Thames Valley Landscapes Monograph 7.
ALLEN, T., BARCLAY, A. & CROMARTY, A.-M. forthcoming. Opening the wood, making the land. The archaeology of a Middle Thames landscape: Mesolithic to earlier Bronze Age. Oxford: Thames Valley Landscapes Monograph.
ALLEN, T., BARCLAY, A. & LAMDIN-WHYMARK, H. 2004. Opening the wood, making the land: landscape in the Dorney area of the Middle Thames Valley. In J. Cotton & D. Field (eds), Towards a new stone age: aspects of the Neolithic in south-east England, 82–98. York: Council for British Archaeology Research Report 137.
ALLEN, T., HEY, G. & MILES, D. 1997. A line of time: approaches to the archaeology of the Upper and Middle Thames Valley, England. World Archaeology 29, 114–29.
BARCLAY, A. 2000. Spatial histories of the Neolithic: a study of the monuments and material culture of central southern England. Unpublished PhD thesis, Department of Archaeology, University of Reading.
BARCLAY, A. 2006. Connections and networks: a wider world and other places. In D. Benson & A. Whittle (eds), Building memories: the Neolithic Cotswold long barrow at Ascott-under-Wychwood, Oxfordshire, 331–44. Oxford: Oxbow.
(p.418) BARCLAY, A. & GLASS, H. with PARRY, C. 1995. Excavation of Neolithic and Bronze Age ring-ditches, Shorncote Quarry, Somerford Keynes, Gloucestershire. Transactions of the Bristol and Gloucestershire Archaeological Society 113, 21–60.
BARCLAY, A. & HEY, G. 1999. Cattle, cursus monuments and the river: the development of ritual and domestic landscapes in the Upper Thames Valley. In A. Barclay & J. Harding (eds), Pathways and ceremonies: the cursus monuments of Britain and Ireland, 67–76. Oxford: Oxbow.
BARCLAY, A., LAMBRICK, G., MOORE, J. & ROBINSON, M. 2003. Lines in the landscape: cursus monuments in the Upper Thames Valley. Oxford: Thames Valley Landscapes Monograph 15.
BARNES, I. & CLEAL, R. M. J. 1995. Neolithic and Bronze Age settlement at Weir Bank Stud Farm, Bray. In I. Barnes, W. A. Boismier, R. M. J. Cleal & M. R. Roberts (eds), Early settlement in Berkshire, 1–51. Salisbury: Trust for Wessex Archaeology, Report 6.
BARTON, N. & ROBERTS, A. 2004. The Mesolithic period in England: current perspectives and new research. In A. Saville (ed.), Mesolithic Scotland and its neighbours, 339–58. Edinburgh: Society of Antiquaries of Scotland.
BATES, M. R. & WHITTAKER, K. 2004. Landscape evolution in the Lower Thames Valley: implications for the archaeology of the earlier Holocene period. In J. Cotton & D. Field (eds), Towards a new stone age: aspects of the Neolithic in south-east England, 50–70. York: Council for British Archaeology Research Report 137.
BAYLISS, A., BRONK RAMSEY, C., VAN DER PLICHT, J. & WHITTLE, A. 2007. Bradshaw and Bayes: towards a timetable for the British Neolithic. Cambridge Journal of Archaeology 17.1, supplement, 1–28.
BAYLISS, A. & HEY, G. in preparation. The scientific dating. In G. Hey (ed.), Yarnton Neolithic and Bronze Age settlement and landscape. Oxford: Thames Valley Landscapes Monograph.
BELL, M. & Walker, M. J. C. 1992. Late Quaternary environmental change: physical and human perspectives. Harlow, Essex: Longman.
BENSON, D. & WHITTLE, A. (eds) 2006. Building memories: the Neolithic Cotswold long barrow at Ascott-under-Wychwood, Oxfordshire. Oxford: Oxbow.
BOISMIER, W. A. & MEPHAM, L. N. 1995. Excavation of a Mesolithic site at Windmill Hill, Nettlebed, Oxon. Oxoniensia 60, 1–19.
BOOTH, P. & SIMMONDS, A. forthcoming. Appleford’s earliest farmers: archaeology at Appleford Sidings, Oxon, 1993–2000. Oxford: Oxford Archaeology Occasional Paper.
BRADLEY, R. 1990. The passage of arms. Cambridge: Cambridge University Press.
BRADLEY, R. 2003. Neolithic expectations. In I. Armit, E. Murphy, E. Nelis & D. Simpson (eds), Neolithic settlement in Ireland and western Britain, 218–22. Oxford: Oxbow.
BRADLEY, R. & GORDON, K. 1988. Human skulls from the River Thames: their dating and significance. Antiquity 62, 271–89.
BRADLEY, R., OVER, L., STARTIN, D. W. A. & WENG, R. 1981. The excavation of a Neolithic site at Cannon Hill, Maidenhead, Berkshire. Berkshire Archaeological Journal 68, 5–19.
BRADY, K. & LAMDIN-WHYMARK, H. forthcoming. Excavation of a Neolithic to Roman landscape at Horcott Pit, near Fairford, Gloucestershire, 2003. Transactions of the Bristol and Gloucestershire Archaeological Society.
BROWN, A. 1995. The Mesolithic and later flint artefacts. In T. G. Allen (ed.), Lithics and landscape: archaeological discoveries on the Thames Water pipeline at Gatehampton Farm, Goring, Oxfordshire 1985–92, 65–84. Oxford: Thames Valley Landscapes Monograph 7.
CLARK, J. G. D. & RANKINE, W. F. 1939. Excavations at Farnham, Surrey (1937–38). Proceedings of the Prehistoric Society 5, 61–118.
CLARKE, D. L. 1976. Mesolithic Europe: the economic basis. In G. de G. Sievking, I. H. Longworth & K. E. Wilson (eds), Problems in economic and social archaeology, 449–81. London: Duckworth.
(p.419) COPLEY, M. S., BERSTAN, R., DUDD, S. N., DOCHERTY, G., MUKHERJEE, A. J., STRAKER, V., PAYNE, S. & EVERSHED, R. P. 2003. Direct chemical evidence for widespread dairying in prehistoric Britain. Proceedings of the National Academy of Sciences of the United States of America 100, 1524–9.
COTTON, J. & FIELD, D. (eds) 2004. Towards a new stone age: aspects of the Neolithic in southeast England. York: Council for British Archaeology Research Report 137.
DARVILL, T. 1996. Neolithic buildings in England, Wales and the Isle of Man. In T. Darvill & J. Thomas (eds), Neolithic houses in northwest Europe and beyond, 77–112. Oxford: Oxbow.
DARVILL, T. 2004. Long barrows of the Cotswolds and surrounding areas. Stroud: Tempus.
DARVILL, T. & THOMAS, J. 1996. Neolithic houses in northwest Europe and beyond. Oxford: Oxbow.
DAY, S. P. 1991. Post-glacial vegetational history of the Oxford region. New Phytologist. 119, 445–70.
ELLABY, R. 2004. Food for thought: a late Mesolithic site at Charlwood, Surrey. In J. Cotton & D. Field (eds), Towards a new stone age: aspects of the Neolithic in south-east England, 12–23. York: Council for British Archaeology Research Report 137.
EVANS, C., POLLARD, P. & KNIGHT, M. 1999. Life in woods: tree-throws, ‘settlement’ and forest cognition. Oxford Journal of Archaeology 18, 241–54.
EVANS, J. G., LIMBREY, S. & MACPHAIL, R. 2006. The environmental setting. In D. Benson & A. Whittle (eds), Building memories: the Neolithic Cotswold long barrow at Ascott-under-Wychwood, Oxfordshire, 55–77 Oxford: Oxbow.
FIELD, D. 1989. Tranchet tools and Thames picks: Mesolithic core tools from the west London Thames. Transactions of the London and Middlesex Archaeological Society 40, 1–25.
FIELD, D. 2004. Sacred geographies in the Neolithic of south-east England. In J. Cotton & D. Field (eds), Towards a new stone age: aspects of the Neolithic in south-east England, 154–63. York: Council for British Archaeology Research Report 137.
FORD, S. 1987. East Berkshire archaeological survey. Reading: Berkshire County Council Department of Highways and Planning Occasional Paper 1.
FORD, S. with ENTWISTLE, R. & TAYLOR, K. 2003. Excavations at Cippenham, Slough, Berkshire, 1995–7. Reading: Thames Valley Archaeological Services Monograph 3.
FORD, S. & TAYLOR, K. 2004. Neolithic occupation at Cippenham, Slough, Berkshire. In J. Cotton & D. Field (eds), Towards a new stone age: aspects of the Neolithic in south-east England, 99–104. York: Council for British Archaeology Research Report 137.
FROOM, F. R. 1971–2. A Mesolithic site at Wawcott, Kintbury. Berkshire Archaeological Journal 66, 23–44.
GLAAS 2004. Greater London Archaeology Advisory Service Annual Review April 2003–March 2004, 7. London: English Heritage.
GRACE, R. 1992. Use wear analysis. In F. Healy, M. Heaton & S. J. Lobb, ‘Excavations of a Mesolithic site at Thatcham, Berkshire’, 53–63. Proceedings of the Prehistoric Society 58, 41–76.
GRIGSON, C. 1989. The animal remains. In B. Stainton, ‘Excavation of an early prehistoric site at Stratford’s Yard, Chesham’. Records of Buckinghamshire 31, 49–74.
HAUGHEY, F. 2000. The Mesolithic on the tidal Thames—a proleptic viewpoint. In R. Young (ed.), Mesolithic lifeways: current research in Britain and Ireland, 221–31. Leicester: Leicester Archaeology Monographs 7.
HAYDEN, C. forthcoming. The prehistoriclandscapeatWhiteHorse Stone, Boxley, Kent. CTRL Integrated Site Report Series, Archaeology Data Service, http://ads.ahds.ac.uk/catalogue projArch.
HEALY, F. 1988. The Anglo-Saxon cemetery at Spong Hill, North Elmham, part VI: occupation during the seventh to second millennia BC. Gressenhall: East Anglian Archaeology 39.
HEALY, F., HEATON, M. & LOBB, S. J. 1992. Excavations of a Mesolithic site at Thatcham, Berkshire. Proceedings of the Prehistoric Society 58, 41–76.
HEY, G. (ed.) in preparation. Yarnton Neolithic and Bronze Age settlement and landscape. Oxford: Thames Valley Landscapes Monograph.
HEY, G. & BELL, C. 1997. Yarnton floodplain 1996. South Midlands Archaeology 27, 62–4.
HEY, G., MULVILLE, J. & ROBINSON, M. 2003. Diet and culture in southern Britain: the evidence from Yarnton. In M. Parker Pearson (ed.), Food, culture and identity in the Neolithic and early Bronze Age, 79–88. Oxford: British Archaeological Reports International Series 1117.
HEY, G. & ROBINSON, M. in preparation. The landscape overview. In G. Hey (ed.), Yarnton Neolithic and Bronze Age settlement and landscape. Oxford: Thames Valley Landscapes Monograph.
HOLGATE, R. 1988. Neolithic settlement of the Thames basin. Oxford: British Archaeological Reports British Series 194.
HOLGATE, R. 2004a. Flintwork. In G. Lambrick & T. Allen (eds), Gravelly Guy, Stanton Harcourt, Oxfordshire: the development of a prehistoric and Romano-British community, 93–9. Oxford: Thames Valley Landscapes Monograph 21.
HOLGATE, R. 2004b. Managing change: the Mesolithic-Neolithic transition in south-east England. In J. Cotton & D. Field (eds), Towards a new stone age: aspects of the Neolithic in south-east England, 24–8. York: Council for British Archaeology Research Report 137.
LAMBRICK, G. 1988. The Rollright Stones: megaliths, monuments and settlement in the prehistoric landscape. London: English Heritage.
LAMBRICK, G. & ALLEN, T. 2004. Gravelly Guy, Stanton Harcourt, Oxfordshire: the development of a prehistoric and Romano-British community. Oxford: Thames Valley Landscapes Monograph 21.
LAWS, G. 2004. Cotswold community, Wiltshire and Gloucestershire: interim report. Oxford: Oxford Archaeology client report.
LEWIS, J. S. C. & WELSH, K. 2004. Perry Oaks—Neolithic inhabitation of a west London landscape. In J. Cotton & D. Field (eds), Towards a new stone age: aspects of the Neolithic in south-east England, 105–9. York: Council for British Archaeology Research Report 137.
LOGUE, P. 2003. Excavations at Thornhill, Co. Londonderry. In I. Armit, E. Murphy, E. Nelis & D. Simpson (eds), Neolithic settlement in Ireland and western Britain, 149–55. Oxford: Oxbow.
MEADOWS, J., BARCLAY, A. & BAYLISS, A. 2007. A short passage of time: the dating of the Hazleton long cairn revisited. Cambridge Journal of Archaeology 17.1, supplement, 45–64.
MITHEN, S. 1999. Hunter-gatherers of the Mesolithic. In J. Hunter & I. Ralston (eds), The archaeology of Britain, 35–57. London: Routledge.
MOFFETT, L., ROBINSON, M. A. & STRAKER, V. 1989. Cereals, fruit and nuts: charred plant remains from Neolithic sites in England and Wales and the Neolithic economy. In A. Milles, D. Williams & N. Gardner (eds), The beginnings of agriculture, 243–61. Oxford: British Archaeological Reports International Series 496.
NEAL, D., WARDLE, A. & HUNN, J. 1990. Excavation of the Iron Age, Roman and medieval settlement at Gorhambury, St Albans. London: English Heritage Archaeological Report 14.
NEEDHAM, S. P. 2000. The passage of the Thames: Holocene environment and settlement at Runnymede. London: British Museum Press.
O’MALLEY, M. & JACOBI, R. M. 1978. The excavation of a Mesolithic occupation site at Broom Hill, Braishfield, Hampshire 1971–1973. Rescue Archaeology in Hampshire 4, 16–38.
O’NEAL, H. 1966. Sale’s Lot Long Barrow, Withington, Gloucestershire 1962–65. Transactions of the Bristol and Gloucestershire Archaeological Society 85, 5–35.
PARKER, A. G. 1995. Late Quaternary environmental change in the upper Thames basin, central-southern England. Unpublished DPhil thesis, University of Oxford.
(p.421) PARKER, A. 1999. Pollen and sediments of Daisy Banks Fen. In A. Barclay & C. Halpin (eds), Excavations at Barrow Hills, Radley, Oxfordshire, Volume 1: the Neolithic and Bronze Age monument complex, 254–67. Oxford: Thames Valley Landscapes Monograph 11.
PELLING, R. forthcoming. The Neolithic and Bronze Age charred plant remains. In T. Allen, A. Barclay & A-M. Cromarty (eds), Opening the wood, making the land. The archaeology of a Middle Thames landscape: Mesolithic to earlier Bronze Age. Oxford: Thames Valley Landscapes Monograph.
PINE, J. & FORD, S. 2003. Excavation of Neolithic, Late Bronze Age, Early Iron Age and Early Saxon features at St Helen’s Avenue, Benson, Oxfordshire. Oxoniensia 68, 131–78.
POLLARD, J. 1999. ‘These places have their moments’: thoughts on settlement practices in the British Neolithic. In J. Brück & M. Goodman (eds), Making places in the prehistoric world: themes in settlement archaeology, 76–93. London: UCL Press.
POLLARD, J. 2005. Memory, monuments and middens in the Neolithic landscape. In G. Brown, D. Field & D. McOmish (eds), The Avebury landscape: aspects of the field archaeology of the Marlborough Downs, 103–14. Oxford: Oxbow.
PRYOR, F. 1974. Excavations at Fengate, Peterborough, England: the first report. Toronto: Royal Ontario Museum Archaeology Monograph 3.
ROBINSON, M. 1992. Environmental archaeology of the river gravels: past achievements and future directions. In M. Fulford & E. Nichols (eds), Developing landscapes of lowland Britain: the archaeology of the British gravels, a review, 47–62. London: Society of Antiquaries Occasional Papers 14.
ROBINSON, M. 1993. The pre-Iron Age environment and finds. In T. G. Allen & M. A. Robinson (eds), Mingies Ditch, Hardwick-with-Yelford, Oxon, 7–19. Oxford: Thames Valley Landscapes Monograph 2.
ROBINSON, M. 1999. The prehistoric environmental sequence of the Barrows Hill area. In A. Barclay & C. Halpin (eds), Excavations at Barrow Hills, Radley, Oxfordshire, Volume 1: the Neolithic and Bronze Age monument complex, 269–74. Oxford: Thames Valley Landscapes Monograph 11.
ROBINSON, M. A. 2000a. Coleopteran evidence for the Elm Decline, Neolithic activity in woodland, clearance and use of the landscape. In A. S. Fairbairn (ed.), Plants in Neolithic Britain and beyond, 27–36. Oxford: Oxbow.
ROBINSON, M. A. 2000b. Further consideration of Neolithic charred cereals, fruit and nuts. In A. S. Fairbairn (ed.), Plants in Neolithic Britain and beyond, 85–90. Oxford: Oxbow.
SAVILLE, A. 1990. Hazleton North: the excavation of a Neolithic long cairn of the Cotswold-Severn group. London: English Heritage.
SCAIFE, R. 1992. Plant macrofossils. In F. Healy, M. Heaton & S. J. Lobb, Excavations of a Mesolithic site at Thatcham, Berkshire, 64–7. Proceedings of the Prehistoric Society 58, 41–76.
SCHULTING, R. J. 2000. New AMS dates from the Lambourn long barrow and the question of the earliest Neolithic in southern England: repacking the Neolithic package. Oxford Journal of Archaeology 19, 25–35.
SERJEANTSON, D. 2003. Review of environmental archaeology in southern Britain. Animal bones. Neolithic and Early Bronze Age (4000 BC–1500 BC): the development of agriculture and animal husbandry. University of Southampton and English Heritage Centre for Archaeology, Unpublished report.
SIDDELL, J. & WILKINSON, K. 2004. The central London Thames: Neolithic river development and floodplain archaeology. In J. Cotton & D. Field (eds), Towards a new stone age: aspects of the Neolithic in south-east England, 38–49. York: Council for British Archaeology Research Report 137.
SPIKINS, P. 2000. Ethno-facts or ethno-fiction? Searching for the structure of settlement patterns. In R. Young (ed.), Mesolithic lifeways: current research in Britain and Ireland, 105–18. Leicester: Leicester Archaeology Monographs 7.
THOMAS, J. 1996. Neolithic houses in mainland Britain and Ireland—a sceptical view. In T. Darvill & J. Thomas (eds), Neolithic houses in northwest Europe and beyond, 1–12. Oxford: Oxbow.
TIMBY, J., STANSBIE, D., NORTON, A. & WELSH, K. 2005. Excavations along the Newbury Reinforcement Pipeline: Iron Age-Roman activity and a Neolithic pit group. Oxoniensia 70, 203–307.
TIPPING, R. 2004. Interpretative issues concerning the driving forces of vegetation change in the Early Holocene of the British Isles. In A. Saville (ed.), Mesolithic Scotland and its neighbours, 45–53. Edinburgh: Society of Antiquaries of Scotland.
WHITTLE, A. 1990. A model for the Mesolithic-Neolithic transition in the upper Kennet valley, north Wiltshire. Proceedings of the Prehistoric Society 56, 101–10.
WHITTLE, A. 1996. Europe in the Neolithic. Cambridge: Cambridge University Press.
WHITTLE, A. 2003. The archaeology of people: dimensions of Neolithic life. London: Routledge.
WHITTLE, A., BARCLAY, A., BAYLISS, A., MCFADYEN, L., SCHULTING, R. & WYSOCKI, M. 2007. Building for the dead: events, processes and changing worldviews from the 38th to the 34th centuries cal BC in southern Britain. Cambridge Archaeological Journal 17.1, supplement, 123–47.
WHITTLE, A., POLLARD, J. & GRIGSON, C. 1999. The harmony of symbols: the Windmill Hill causewayed enclosure, Wiltshire. Oxford: Oxbow.
WILLIAMS, E. 1989. Dating the introduction of food production into Britain and Ireland. Antiquity 63, 510–21.
ZVELEBIL, M. 1994. Plant use in the Mesolithic and its role in the transition to farming. Proceedings of the Prehistoric Society 60, 35–74.
Proceedings of the British Academy 144, 399–422, © The British Academy 2007.