Regional Perspectives on Neolithic Pit Deposition: Beyond the Mundane

By Julian Thomas

The rise to prominence of pits within narratives of the British and Irish Neolithic is well-documented in recent literature. Pits have been cropping up in excavations for centuries, resulting in a very broad spectrum of interpretations but three main factors have led to the recent change in our perception and representation of these features: a broad shift in people's expectations as to what a Neolithic settlement should be; the development of the concept of 'structured deposition', within which pits have played a key role; and a dramatic rise in the number of pits actually known about. Development-led archaeology, and the often very large areas its excavations expose, has simply revealed many more pits. The 15 papers in this volume explore these inter-related factors and present new thoughts and interpretations arising from new analysis of Neolithic pits and their contents.

• Language: English
• Publisher: Oxbow Books
• Release date: Dec 14, 2011
• ISBN: 9781842177051

Book preview

Chapter 1

Introduction: beyond the mundane?

Julian Thomas

THE HUMBLE PIT

Pits containing cultural debris are one of the more common features of the Neolithic archaeology of Britain, and over the past century they have attracted the attention of successive generations of archaeologists for a variety of different reasons. In the earlier twentieth century, pits were recognised as ‘closed finds’, within which valuable sealed assemblages and associations might be identified (Smith 1910), or occasionally as dwellings in their own right (Stone 1934). As the numbers of known pit sites increased, providing the raw material for Isobel Smith’s (1956) momentous study of Neolithic pottery in south-east England, a different approach to the phenomenon began to emerge. In a singularly important paper, Field, Matthews and Smith (1964, 367) suggested that during the Neolithic houses were as a rule constructed in such a way as to leave no permanent or recognizable traces in the ground. In consequence, pits that had been used for storage were likely to represent the surviving component of Neolithic settlements. Noting that pits of Neolithic date appeared to be preferentially distributed to the south and east of the Jurassic Ridge, Field, Matthews and Smith hypothesised an upland/lowland economic divide modelled on Stuart Piggott’s Iron Age ‘Woodbury’ and ‘Stanwick’ regimes (Piggott 1958). That is to say, the presence of pits used for storage was an indicator of an arable economy, while their absence suggested pastoralism. However, recent years have seen the distribution of Neolithic pits extend into Scotland, Wales and Northumbria, as well as many parts of Ireland, eroding the force of this argument (Brophy and Noble, Chapter 6).

Five years later, Humphrey Case described the results of his excavation at the enclosed site of Goodland in Antrim, which contained numerous intercutting pits. These had phosphate-rich fills, and many contained clusters of potsherds, flint flakes, fragments of burnt bone, and animal teeth (Case 1973, 188). Noting the similarity between these deposits and the chamber fills of many court tombs, Case proposed the striking interpretation that this material had been scooped up from rubbish scatters on abandoned settlements and conserved owing to its perceived fertility. The presence of sherds and flints would have served as indicators of enriched soils where healthy crops would grow, and Case speculated that such material would have been selected for use in sympathetic magic. Alongside this recognition that pit contents (and other deposits) might have been deliberately placed we can identify a growing concern with assemblage variation, and the notion that burial might represent a means by which fine artefacts could be taken out of circulation (Bradley 1982; 1984, 51). The concept of ‘structured deposition’ emerged from these concerns, identifying elements of the archaeological record that appeared to have been generated purposively, in a formal and perhaps rule-bound fashion. Whilst these ideas were developed in order to address the rather unusual assemblages contained within the henge monument of Durrington Walls in Wiltshire, it was suggested that some degree of formality could be detected in other kinds of deposits, such as pit-fills (Richards and Thomas 1984, 215). This possibility was later enlarged upon by considering pits in the context of a broader suite of Neolithic depositional practices. Being generally small and bowl-shaped rather than large and cylindrical, Neolithic pits are not well suited to the storage of plant foods in a damp climate. They generally show little sign of weathering or silting before their swift backfilling. This prompts the suggestion that they were created for the purpose of deposition, without having had any previous use. Rather than a slow build-up of successive layers, Neolithic pits often have a single homogeneous fill, or a small number of fills that were deliberately deposited in quick succession. These fills very often contain burnt material, whether charcoal, ash, or hearth debris (Thomas 1999, 65–6).

One of the features of Neolithic pits that originally suggested that they might represent something other than storage facilities or recipients of casual discard was the presence of high proportions of stone tools as opposed to waste flakes, and of especially fine or attractive objects, which had sometimes apparently been carefully arranged within the space of the pit. A case in point would be the group of Grooved Ware pits close to the Dorset Cursus in Firtree Field, Down Farm. Here, pit contents included arrowheads, polished flint implements, boar’s tusks, a complete ox skull, large slabs of pottery, and stone axes (Barrett, Bradley and Green 1991, 77). While a focus on these spectacular discoveries was important in establishing the idea that Neolithic pit deposits constituted a distinct class of evidence, and were worthy of study in their own right, the consequence could be a neglect of the less impressive majority of pits, as a number of contributors to this volume maintain (see Becket and MacGregor, Chapter 5; Edwards, Chapter 7). The recent growth of interest in Neolithic pits has coincided with an explosion of the quantity of information available about them, much of it generated in developer-funded archaeology. Yet the relative humility of many pits and their contents has often resulted in their relegation to the grey literature. This is an unsatisfactory state of affairs, since it is the very ubiquity of Neolithic pits that makes them an invaluable indicator of social, cultural and economic variability over space and time. While for some prehistorians they are just inscrutable pits (Rowley-Conwy 2003, 124), they arguably offer a key measure of regional diversity in the Neolithic.

It was with this point in mind that Hugo Anderson-Whymark and the author proposed a meeting on pit deposition to the Neolithic Studies Group, which was held at the British Museum in November 2009. We assembled a group of speakers who could address material from different parts of Britain and Ireland, and asked them to consider a series of aspects of variation in practice, as a basis for comparison. These included the process of deposition itself, the contents of pits, relationships between pits, changes over time, and relationships with other kinds of sites and structures (see Anderson-Whymark, Chapter 13). In the event, while subtle and fascinating contrasts between regions did emerge, it was the set of regularities spanning Britain and Ireland that struck many of those who attended the meeting. In particular, the incorporation of occupation material into pit fills, including hearth material that may be redolent of human sociality and inhabitation, is extremely widespread. In the case of early Neolithic pits this debris has often been poured into the cut features in a relatively undifferentiated state, but there are indications of more careful filling and a greater degree of selection of the objects included as the period progressed. The clear link between pit-digging and places of occupation sets up a series of themes for investigation: the role of pits in memory and commemoration; the connection between people and specific places in the landscape; the rhythms of everyday life, and the ways in which people give structure to their own existence.

Pit deposits appear to amount to a coherent and long-lived tradition of practice in Neolithic Britain and Ireland, but quite how this custom of physically introducing the traces of human habitation into the landscape first became established is a surprisingly difficult question to answer. Darvill (Chapter 3) draws attention to the elaboration of natural solution features at Billown on the Isle of Man, a phenomenon also observed at Eaton Heath in Norfolk, Cannon Hill in Berkshire, and the Firtree Field doline in Dorset (Wainwright 1973; Bradley, Startin, Over and Weng 1978; Green and Allen 1997). This might conceivably have built upon a concern with ‘natural places’ that extended back into the Mesolithic (Bradley 2000, 35). Equally, Mesolithic pits containing placed deposits are known both in Britain and on the continent (Darvill, Chapter 8; Kirk 1991, 123), while Lamdin-Whymark points out that there is continuity in the practice of depositing cultural material in tree-throw holes across the boundary between Mesolithic and Neolithic (2008, 137). But on the other hand, pit-digging and filling was a characteristic feature of the Neolithic in many parts of Europe, often taking on a distinctly structured character (e.g. Chapman 2000, 69). More pertinently, bowl-shaped pits containing deliberately-placed occupation debris are a feature of the Early Neolithic in southern Scandinavia, where they often define inhabitation sites that are distinct from the contemporary house-sites, and perhaps characteristic of more transient forms of habitation (Anderson, Karsten, Knarrström and Svensson 2004, 159; Larsson and Rzepecki 2003, 7).

PITS AND OCCUPATION

While the material contained within pits is often apparently related to domestic habitation, more direct connections with dwelling (or dwellings) are particularly informative. At the recently-excavated late Neolithic settlement at the eastern entrance of Durrington Walls, the construction of each house was preceded by a series of borrow pits from which material was extracted for wall daub and floor plaster. Each house also had a ‘decommissioning pit’, which often cut through the floor of the building. These latter contained large quantities of pottery, animal bones, and stone tools including arrowheads, and are comparable with rich Grooved Ware pits reported from elsewhere in Britain (Parker Pearson 2007, 138). So effectively the life-history of the house was bracketed between two episodes of pit-digging, and we could say that pit episodes segmented time and correlated with critical events in the life of a household. This is an insight that may be of broader significance. At the two Scottish early Neolithic halls of Claish and Warren Field, a large proportion of the pottery assemblage from each site was recovered from substantial pits located on the axis of the building. The fills of each pit included burnt material, and artefacts appeared to have been deliberately positioned. At Warren Field the pits concerned had previously held timber uprights that had been withdrawn before the burning of the hall, and the same was probably the case at Claish (Murray, Murray and Fraser 2009, 38; Barclay, Brophy and MacGregor 2002, 77). These posts were apparently not structural, and may have had a social importance, resulting in their removal as part if the closing-down of the hall. So here again, deliberate deposition was intimately connected with the termination of activity at a building, part of whose use might have been as a dwelling, for at least part of the year. Several of the chapters in this book deal with the relationship between pits and houses or halls. Smyth (Chapter 2) notes that there are pits associated with early Neolithic houses in Ireland, but that there are also contemporary sites with pits but without substantial above-ground structures. At the site of Green on Eday, Coles (Chapter 4) describes pits in the context of a substantial stone-built house. At Horton in Berkshire, by contrast, Chaffey and Brook (Chapter 14) show that a group of pits was rather later in date than the timber house, and probably relates to a different kind of occupation of the site, although they speculate that the ‘void’ between the pits might have been occupied by an ephemeral structure.

Some while before the excavation at Durrington east entrance, Joshua Pollard had already suggested that pits might represent formal statements of abandonment, gathering up material strewn across the surface of a short-lived settlement (Pollard 2001, 323). However, the detailed analysis of artefacts from pit-groups at Kilverstone in Norfolk indicates that each series of intercutting features may have been created at intervals during the course of a single episode of inhabitation (Garrow, Beadsmoore and Knight 2005, 152). Potentially this might mean that there were circumstances other than the commencement and termination of occupation that might trigger pit deposition. In many of the areas discussed in this book, concentrations of Neolithic pits were located on slight eminences overlying relatively flat, low-lying country (Smyth, Chapter 2; Jackson and Ray, Chapter 11; Garrow 2007, 9). While in the Thames valley and the Severn-Wye region pits typically occurred in pairs or small clusters, very large pit groups are a notable feature of the Neolithic in East Anglia (Anderson-Whymark, Chapter 13). The latter include Broom Heath, Hurst Fen, Kilverstone and Spong Hill, all of early Neolithic date and composed of dozens of pits (Garrow 2007, 3). At each of these sites, the objects contained within the pit appear to have been introduced to the feature together with the fill matrix, and without having been deliberately arranged to any extent. All of the pottery vessels involved were fragmentary, and the lithic material is dominated by waste flakes. The artefacts show extensive signs of weathering and burning, and there appears to have been a significant interval between the accumulation of the cultural material in a ‘pre-pit context’ of some kind and its final deposition (Garrow 2007, 12). The formation of early Neolithic pits in Ireland appears to have been very similar to this, although they generally occur in smaller numbers (Smyth, Chapter 2). Much the same could be said of early Neolithic pits in the Thames valley, with artefacts ‘jumbled through the matrix’ (Anderson-Whymark, Chapter 13), although in the case of a series of pits containing primary Neolithic Carinated Bowl sherds from Ayrshire, Becket and MacGregor (Chapter 5) described the ‘mixing’ of artefacts into the pit fill. This would seem to imply that objects were purposefully added to the charcoal-rich soil, as Edwards (Chapter 7) argues for Thirlings, where sherds of a particular size appear to have been selected for inclusion.

Although the East Anglian pit sites were extensive, they need not have resulted from the activities of exceptionally large communities. At Kilverstone, each of the distinct clusters of pits within the site appears to have drawn on material from a separate pre-pit context (Garrow, Beadsmoore and Knight 2005, 150). Beadsmoore, Garrow and Knight argue that the lack of refitting sherds and flints between spatially separate groups of pits is a consequence of their temporal separation: each relates to a distinct episode of occupation (2010, 123). In a similar fashion, Jackson and Ray (Chapter 11) point to refits between pairs of pits in the Severn-Wye region, indicating that these are probably also contemporary. One of the important points demonstrated by Garrow, Beadsmoore and Knight’s meticulous work at Kilverstone is that while the filling of each pit would have taken place in a short period of time, this would have cut across a series of processes whose temporalities were quite different (2005, 151). The filled pit is a stable context within which a series of artefact biographies terminate and are ‘bundled together’. But the objects and materials contained within the pit would previously have been caught up in quite different cycles and rhythms: the lithic reduction sequence, and the use and discard of stone tools; the use, breakage, wear and burning of pots; the accumulation of surface dumps and deposits, and the rotting and decay of their organic component; the burning and raking-out of fires; the seasonal or longer-term movements of the human community, and its episodes of fission and fusion. The pit fill is effectively a cross-section of these intersecting temporal pathways, caught at a particular moment. The important methodological innovations of the Kilverstone study demonstrate how these complicated processes can be unpicked, and in this volume Emilie Sibbesson (Chapter 9) shows that the approach can be pushed still further, by investigating the fabrics of the pottery vessels.

Pollard (2001, 323) made the observation that the residues of habitation might not have been understood simply as ‘waste’ by Neolithic people, so much as matter that had gained potency or efficacy through its involvement in social life. In this he echoes Case, identifying settlement debris as material that needed to be treated with respect and caution. There is possibly a comparison to be made with antler picks used in the construction of Neolithic monuments, and often left lying in ditches and postholes (for instance, Wainwright and Longworth 1971, 22). The implication is that their role was more than a purely instrumental one, and that their removal for further use elsewhere might have been inappropriate. In the same way, the burial of occupation residues may have been guided by the conviction that they amounted to more than inert matter. When people placed the residue of their everyday activity in pits, perhaps before moving on, it is conceivable that some connection was being asserted between the community and a specific location. To think of this in terms of ‘territoriality’ might be a little crude. Jackson and Ray (Chapter 11) make reference instead to the notion of ‘presencing’, in which material things metaphorically evoke absent persons with whom they are associated. We could perhaps connect this argument with Nancy Munn’s discussion with the way in which people in non-state societies may seek to extend their own presence across space and time through the circulation of artefacts, thereby securing their widespread recognition, fame or renown (Munn 1986, 9). If the physical traces of specific people were acknowledged as being embedded in particular locations, they might achieve the goal of being brought to mind in their absence.

RITUAL, BACKGROUND AND ‘MATERIAL STATEMENTS’

If ‘structured deposition’ was originally identified in the context of a discussion of ritual action (Richards and Thomas 1984, 190), one unfortunate consequence has been that Neolithic features with elaborate fills have sometimes been identified as ‘ritual pits’. With good reason, there has been a recent reaction against the attempt to identify ritual and the domestic as opposed and mutually exclusive categories (Bradley 2005, 28). Joanna Brück (1999, 314) points out that this distinction is sometimes extended to an opposition between the practical and the symbolic, or the sacred and the secular. As she goes on to say, ritual activity is itself highly practical, since it is often intended to achieve objective ends in the world, through the influence of metaphysical agencies. Consequentially, it will not be easy to uncouple symbolic and pragmatic aspects of human action. While agreeing with all this, we need not necessarily follow Brück to the conclusion that archaeologists should dispense with the concept of ritual altogether, investigating instead logics of action different from our own (Brück 1999, 336). The key point is to recognise that ritual is not a separate sphere of action, but a mode of conduct that people slip into and out of in the course of their everyday lives. Ritual does not necessarily employ a different logic from domestic life, although it may provide a context in which that logic comes more sharply into focus. Ritual can be defined as expressive performance that is prescribed by appropriateness, and is sanctioned by tradition (Lewis 1980, 8). Features such as the formality of movement and body posture, or the archaic character of utterances, serve to alert people that ‘something different’ is going on, resulting perhaps in an enhancement of attentiveness and a heightening of sensation. Yet this ‘something different’ may be no more than saying a prayer before a meal, offering a blessing before slaughtering an animal, or washing in a prescribed manner in preparation for a special event. There is therefore a continuum of ritualised acts, only the most elaborate and important of which could be distinguished as ‘ceremonial’. And only a minority would make use of regalia, paraphernalia and structures that were made with an exclusively ritual use in mind. Although ritual makes use of symbols as its fundamental elements, most of these symbols are drawn from everyday life (Turner 1969: 15). So while we might argue that it is appropriate to consider some prehistoric monuments as ‘ritual structures’ (and even this need not have been the case), it is less helpful to talk of ‘ritual pits’. Instead, we may be dealing with pits whose filling with everyday materials involved a degree of ritual action.

In a majority of cases the level of ‘ritual’ concerned will have been no more than following a traditional practice in an approved manner, perhaps with rather little deliberation at all. Yet this established way of doing things provided the background against which more elaborate deposits and more emphatic ‘material statements’ could stand out. The rather undistinguished character of early Neolithic pit fills in East Anglia, where Garrow (2007, 16) notes that ‘special objects’ are entirely lacking, seems to have been the norm throughout much of Britain and Ireland. Yet there are striking exceptions. Cannon Hill, Roughridge Hill and the Coneybury ‘Anomaly’ were all large pits with very rich material assemblages. Pollard (2001, 323) notes that at the latter two sites the objects had been carefully arranged into groups and clusters, and that someone must actually have entered the pit in order to achieve this. All three sites date to the earliest centuries of the Neolithic, before the construction of causewayed enclosures, and Cleal (2004, 176) suggests that they may have been the product of episodes of large-scale gathering and consumption of food. Certainly, the large number of animals butchered at Coneybury implies more than the routine culinary activity of a small domestic unit. Indeed, Reynolds (Chapter 12) makes the argument that what occurred at Coneybury was not merely a feast, but an event whose significance drew upon its specific location. Another potentially ‘special’ early Neolithic pit site has recently been reported at Hatton Farm, Angus, where sherds from no less than 132 Carinated Bowl vessels were crammed into a shallow oval pit, together with a mass of burnt sandstone cobbles (Gray and Suddaby 2010, 8). If these deposits really did result from comparatively large social gatherings, their relationship with ‘occupation’ pits remains to be fully understood. At Coneybury at least a proportion of the material had probably lain in a midden for at least a short while before deposition, so the pit may not have simply been filled in the process of ‘tidying up’ in the immediate aftermath of a feast (Richards 1990, 42).

If only a small minority of early Neolithic pits reveal this kind of complexity, the papers in this book indicate a general trend toward greater formality in Later Neolithic deposits, and especially those associated with Grooved Ware. Certain classes of material were more likely to be included in Grooved Ware pits, including human remains (Pannett, Chapter 10), and stone axes such as the group from Clifton reported by Jackson and Ray (Chapter 11). Objects in Grooved Ware pits are also more likely to have been affected by fire. As well as artefacts, even plant remains in late Neolithic deposits had sometimes been carefully selected (Chaffey and Brook, Chapter 14). Pollard (2001, 325) noted that Grooved Ware pits in southern England were generally pleasingly sculpted, with smoothly rounded bases, and that their contents had often been arranged before the fill had been introduced. The implication of this is that it was not enough for objects and substances to be put into the ground: the process of deposition had to be witnessed by people other than the pit-digger, rather like the mourners gathered beside a grave. Ben Edwards (Chapter 7) suggests a rather similar development when he contrasts the undifferentiated pit fills at Thirlings with the visually arresting deposits inside the Milfield North and South henges. Layering, ordering and formal arrangements of objects are also described in Ireland and the Isle of Man (Smyth, Chapter 2; Darvill, Chapter 3), although Garrow (2007, 14) suggests that the pattern may be less emphatic in East Anglia than elsewhere in southern Britain, with the ‘stack’ of scrapers in a Grooved Ware pit at Over a comparatively rare instance of formal arrangement.

PITS, MEMORY AND MIDDENS

Part of the significance of the sporadic burial of settlement residues would have lain in engendering memory (Pollard 2001, 323). At numerous sites including the Coneybury, Flagstones, Balfarg, and Pict’s Knowe enclosures, and the Holywood South cursus, monuments were built on sites that had previously been used for pit deposition (Thomas 2007, 263). Having been identified as important places, these locations retained their importance over appreciable periods of time. It is arguable that the pit-deposits were merely the by-product of some more important happening, and that this was what was principally being remembered, but it remains the case that digging and deposition must have had a place in giving a tangible quality to the process of remembering. Conversely, during the later Neolithic pits were often dug into what were by then quite ancient structures, such as the timber halls at Yarnton and White Horse Stone, and the causewayed enclosure at Etton (Hey, Mulville and Robinson 2003, 81; Pryor 1998, 111). In these cases pit deposition perhaps amounted to a form of commemoration, or even veneration.

Another way in which pits were potentially connected with themes of memory and connection with the past lay in the incorporation of curated artefacts. In discussing the Beaker grave assemblage from Raunds Barrow 1, Healy and Harding (2004, 186) have drawn attention to the way that objects with protracted histories can lend associations with the ancestral past to a funerary deposit. Several of the contribution in this book hint that something similar applied to pit deposits. At Cilsan in Carmarthenshire, Pannett (Chapter 10), describes how sherds of Grooved Ware, Collared Urn and Food Vessel were recovered from a single pit, showing signs of weathering and abrasion indicating that they had been in circulation for some time before deposition, presumably at some point in the early Bronze Age. Equally, Jackson and Ray (Chapter 11) report the presence of sherds of Mortlake Ware from an otherwise Grooved Ware-dominated assemblage at King’s Stanley. Most remarkable of all is the large Peterborough Ware rim sherd from a Grooved Ware pit at Horton, which produced a radiocarbon date several centuries earlier than the remainder of the assemblage (Chaffey and Brook, Chapter 14). Now, all of these objects could conceivably have been thrown into the pits concerned along with fill material drawn from long-standing middens. Yet since this seems to be a variation in practice characteristic of Grooved Ware pits, amongst which deposited artefacts appear to have been selected, it is likely that the comparative antiquity of these objects was appreciated to some degree.

This suggests that the relationship between pits and middens is worthy of consideration. Garrow, Beadsmoore and Knight (2005, 149) inferred from their work at Kilverstone that there must have been a delay between broken pots and knapped flints being generated and their incorporation into pit fills. The ‘post-breakage, pre-pit context’ must have been a place of accumulation, where these materials must have been mixed with hearth ash and organic materials such as food remains, and subjected to variable degrees of weathering, so that sherds from the same vessel acquired differing degrees of wear. The kind of context that they had in mind might have been similar to the surface spreads of midden material revealed at Dorney Lake Areas 6 and 10 (Allen, Barclay and Lamdin-Whymark 2004, 89). These were composed of distinct dumps of material including ash and charcoal, which suggested repeated episodes of occupation and recurrent use of depositional loci. Nearby tree-throw holes had been filled with material drawn from these surface middens, and, significantly, pieces of flint were retrieved from the spreads and reworked. It is to be presumed that deposits like those at Dorney will often not have been preserved, and having been ploughed away they are represented only by flint scatters retained in the topsoil. Interestingly, in the course of his investigation of the Neolithic settlement of the Thames valley, Holgate (1988, 78) observed that some groups of pits were associated with surface flint scatters, while others were isolated. An example of the latter discussed in this volume is the Impressed Ware pit at Brackmont Mill, Fife (Brophy and Noble, Chapter 6). So, were such isolated pits the consequence of short-term and perhaps unrepeated occupation, or do they relate to some other kind of activity? This obviously has some bearing on the ‘special’ pits such as Coneybury and