Fig. 1_. General location plan showing situation of the three major sites discussed in the paper, East Chisenbury, All Cannings Cross and Stanton St Bernard.

THE EXCAVATIONS AT ALL CANNINGS CROSS, WILTSHIRE

1In 1911 Maude Cunnington began the excavations that established All Cannings Cross as the type site for the Early Iron Age of Southern Britain. This contribution will consider how archaeological understanding of this period has changed since those excavations took place and, as a resuit, why we hâve decided to undertake further fieldwork at the site.

2Cunnington’s attention was drawn to All Cannings Cross by the objects of worked stone brought to the surface as the result of ploughing, and by the time her excavations were completed in 1922 (the work was interrupted by the first world war) she had recovered a substantial quantity of pottery, worked bone, metalwork, animal bone and worked stone from the site (Cunnington 1923).

3All Cannings Cross lies on the northem margins of the Vale of Pewsey, a deep eut through the chalk uplands of Southern England separating the Marlborough Downs to the north from Salisbury Plan to the south (fig. 1). These uplands are well explored archaeologically; the great Neolithic monuments of Avebury lie to the north while Stonehenge sits centrally within Salisbury Plain to the south. The archaeological history of the lowland Vale in which All Cannings Cross and Stanton St. Bernard are located is poorly understood, although the area hosts the important Late Neolithic complex at Marden. In addition, there are a number of Iron Age hillforts, as well as two Neolithic causewayed enclosures, immediately overlooking the Vale.

4The importance of the All Cannings Cross material was not immediately apparent to Cumiington. However, in 1915 Bush-Fox published the results of excavations at Hengistbury Head, on the south coast of England. In that publication Bush-Fox (1915) established a cultural sequence for the Iron Age finds, beginning with pottery that he assigned to his « Class A » and which he compared to Early Iron Age Hallstatt material from the Jura. In so doing he pointed towards a continental assemblage that provided both a cultural identity for the earliest Iron Age in Britain, and a historical context for that horizon in terms of its European origin.

5Cunnington was quick to realise the implications of the Hengistbury Head sequence for her own, far more substantial, body of material and by the time she published she was able to lay before the reader a cultural assemblage that evoked the end of one dominant technology (bronze) and the beginning of another (iron). The former was represented at All Cannings Cross by the blade of a Late Bronze Age Armorican socketed axe (Cunnington 1923: pl. 18.3) while the latter was marked by a distinctive séries of iron pins, also with continental parallels (Cunnington 1923: pl. 21.1-5). Bush-Fox’s Class A ceramic tradition was now more clearly defined by the large and well preserved assemblage of decorated sherds from Ail Cannings Cross. And whilst later material clearly existed on the site, including two early La Tène brooches (Cunnington 1923: pl. 18.12-13, pl. 21. 7-9), it was the Early Iron Age Hallstatt C horizon that the All Cannings Cross material came to represent.

6If the All Cannings Cross artefact assemblage was destined to define a well-known horizon in the cultural development of the British Iron Age, the character of the site itself was far from clear (fig. 2). Iron Age settlement studies in Southern Britain were not well developed when Cunnington wrote and the major field monuments of the period, hillforts, were only just starting to be investigated with any consistency. There was, therefore, no obvious context with which to compare the deep ‘humic’deposits that Cunnington encountered, nor did comparanda exist for the rectangular chalk and clay floors and other structural details she had recorded. The post-built structures thus remained unanalysed, missing a possible large post-built round house in the centre of Cunnington’s excavation (Cunnington 1923: pl. 2).

Fig. 2. Cunnington’s excavation plan. She opened her trenches in areas of high densities of surface artefacts. The Sheffield University project suggests that the location and extent of these trenches were not surveyed and drawn with any great accuracy.

CHANGING PERSPECTIVES 1922-1982

7In 1959 Christopher Hawkes published his « Scheme for the British Iron Age ». This marked the culmination of some thirty years of work in which Hawkes and others had assessed the growing body of excavated material from hillforts and settlement sites in Southern and eastern Britain alongside the limited and regionally specific burial evidence of the period. The origins of this cultural sequence of material was set against the then current understanding of Late Bronze Age settlement activity that derived the Bronze Age cremation cemeteries of the Deverel Rimbury Culture from continental Urnfield traditions. Maintaining the ABC nomenclature of Bush-Fox, Hawkes divided the British Iron Age sequence into three continentally derived stages of cultural development. British later prehistory took the form of horizons of continental contact, migration or invasion interspersed with periods of regional development. By this means hillfort and settlement development could be narrated as resulting either from continental impositions, indigenous reaction to continental incursions, or indigenous development. Key assemblages used to support this scheme included the assumed Hallstatt C material from Hengistbury Head and All Cannings Cross, a small group of ceramic finds from Sussex in Southern England and the burials of eastem Yorkshire representing a La Tène, « Marnian » invasion, and the cemeteries at Aylesford and Swarling that were assumed to relate to the arrival of the Belgae, an event that was believed to have been securely attested by Caesar (Hawkes 1959, 1960).

8The timing of Hawkes’publication was inauspicious for the tide had begun to turn against the « invasion hypothesis » as a means of explaining the development of British prehistory. There were two reasons for this, one general, the other more spécifie to Iron Age studies. In general social, economic, and cultural change had corne to be seen to result more from a complex interplay of processes operating within indigenous settlement Systems than from any extemal imposition such as might have resulted from continental migrations. This change in perception drew its strength from the development of economie archaeology as well as from the reaction against diffusionist arguments as they had been applied more widely to European prehistory (cf. Clark 1952, 1957). The combined intellectual force of this change in thinking was behind much of the British contribution to the development in the 1960s of what was then known as the « New Archaeology » (cf. Clarke 1968; Renfrew 1968).

9The second, more spécifie reason for rejecting the scheme proposed by Hawkes was empirical: the evidence simply did not support the interpretation that was being placed upon it. Hodson rehearsed this criticism first by questioning the ceramic typology proposed to support « Marnian » horizon in Southern Britain (Hodson 1962). He then went on to tackle the full Hawkesian scheme by rejecting the suggestion that the cultural sequence of the British Iron Age was to be derived from the continent. Instead he argued that the Iron Age was demonstrably an indigenous development of agricultural communities derived from the Late Bronze Age (Hodson 1964). He labelled this indigenous Iron Age the « Woodbury Culture » after the agricultural settlement of Little Woodbury that had been partially excavated by Bersu in 1938 and 1939, a cultural group to which he was happy to assign the material from Ail Cannings Cross. Any continental material was, according to Hodson, of only marginal significance to the overall development of the British Iron Age.

10Hawkes never provided the promised réfutation of Hodson’s critique and since the 1960s Iron Age studies in Britain hâve focused almost entirely upon tracing continuity in the development of settlement organisation and agricultural practices. In many ways Hodson’s critique was not particularly radical, it was simply a re-writing of a cultural history from an indigenous rather than a continental perspective, and it was left to others to seek to explain the way the social and économic processes of the period gave rise to the changes witnessed in settlement form and organisation. General models of political and économic centralisation in the European Iron Age were developed for example by Collis (1984), whilst in Britain Bradley (1971) sought an indigenous, agricultural origin for hillforts and Cunliffe embarked on the analysis of the Danebury hillfort, a hillfort which he believed developed as a régional économie centre (Cunliffe 1984). It should be noted however that change, when it cornes at the end of the Iron Age, still tends to be explained by the impact of external influences in the form of Mediterranean and Armorican trading contacts along the channel coast-line of Britain (Cunliffe 1987).

CHANGING PERSPECTIVES AFTER 1982

11All Cannings Cross, known primarily as an assemblage of cultural material, and lacking any secure understanding of its settlement organisation and economic activity, would seem to have lost much of its relevance to archaeological research witli the abandonnant from the early 1960s onwards of the cultural and ethnic narratives that sought to explain the development of the Iron Age. The assemblage might still define what we should expect a regional assemblage marking the begiiming of the Iron Age to look like, but it could no longer be used to define a continental Hallstatt C incursion into Britain. Our current understanding of All Cannings Cross, therefore, fails to offer any information that might help to explain the changes that accompanied the adoption of iron technology. Our decision to retum to the site, consequently, requires some explanation: what issues can All Cannings Cross now help us to address?

12The conflicting views of the begimting of the Iron Age, as either resulting from continental cultural influence or from indigenous continuity, seem to offer equally unlikely options. It is unnecessary to revert to the « invasion hypothesis » to recognise the problems associated with Hodson’s model for the period. His argument was that the agricultural landscape of the Southern British Late Bronze Age provided the basis for the developing settlement patterns of the Iron Age. This view has been complicated by three factors: evidence for settlement discontinuity; the implications of technological change; and the evidence for new forms of communal practices at the end of Bronze Age and beginning of the Iron Age.

Fig. 3. This small U-shaped enclosure dates to c. 1500 calBC and both it and the adjacent linear earthwork overlie, and thus decommission, earlier fields. This may well imply that a certain amount of economie re-organisation took place in the middle of the 2nd millennium BC.

13For Hodson the origins of his Woodbury Culture lay in the Bronze Age settlement patterns then assigned to the Deverel Rimbury Culture. These settlements were represented by round-houses, enclosures, associated field-systems and cremation cemeteries, classic examples of which had been excavated on the chalk uplands of Southern Britain (fig. 3). There were however two problems with this derivation. First it was clear by the time Hodson was writing that this material was most securely dated to around 1200BC and therefore to the Middle and not Late Bronze Age (Sntith 1959). Subséquent work has extended this chronology earlier, towards 1700BC in some areas (Framework Archaeology 2006), while a latter currency cannot be sustained, thus breaking the cultural continuity presumed by Hodson. Secondly, the clarity with which the upland seulement evidence had been recorded was entirely due to the fact that these settlements were not overlain by later structures nor eroded by later agricultural activity. They therefore appear more as the latest, and abandoned, stage of seulement expansion than as the spring-board for a continuons sequence of Iron Age settlement.

14As more evidence has accumulated and the nature of Late Bronze Age settlement has become clearer then the indications of quite rapid changes in the organisation of seulement in the period have been recognised. In some areas a pattern of dispersed, relatively small-scale seulement nuclei are abandoned for the development of nucleated and longer lived foci while, as we have seen, some other areas of settlement are abandoned. This re-orientation of seulement organisation is, in part, the basis for the development of the first hillforts. These structures remain more firmly associated with the very beginnings of iron technology than might have been implied by arguments for continuity between the periods. In the agricultural landscape the changes in settlement organisation are also manifest in the abandonment of large areas of upland field Systems and the building of extensive linear earthworks (fig. 4), perhaps demonstrating an increasing emphasis upon the territorial definition of landscape and resources (Bradley et al. 1994).

15It would be wrong to take these broad généralisations too far because régional variations do occur, but the important point is that the Late Bronze Age was a period of considerable economic and social development in Southern Britain and that the adoption of an iron technology should be seen in that context. And this dynamism is unlikely to be attributed to entirely local processes such as population growth or increased agricultural activity. The lowlands of Britain shared, with other lowland regions bordering the North Sea and the Channel, the need to import metal supplies, often at some considerable distance, to maintain bronze production. These local economies were therefore linked if not by large-scale migration then by trade and networks of communication into larger « world economies ». The archaeology of these larger systems is attested by the distribution of metal types, the hoarding and recycling of metal, and by the evidence for maritime trade. By the 1980s the realisation of how important these large scale exchange networks were likely to have been resulted in modelling the kinds of political structures that may have controlled a proportion of the local economies through systems of ritualised authority and the circulation of prestige materials (Rowlands 1980). The strength of these models was that they offered an understanding of the link between local productivity, long-distance exchange and ritual display, and they helped to explain some of the possible ways these links might have generated their own compétitive dynamism and change. The problem was, however, that the intellectual attractiveness of these ideas became self-sustaining; they seemingly required little by way of empirical validation and indeed the archaeological evidence always appeared to fit in with theoretical expectations. However, one important implication of the ideas still remains: the change from a dominant bronze to iron technology in the eighth century BC will have had a dramatic impact on the organisation of long-distance exchange and the dependency of both political authority and agricultural économies upon that trade. Indeed it seems possible that the adoption of iron enabled the expansion of settlement into some more marginalised regions as the dependency on long-distance trade links by the local economies was loosened.

16The inevitable conclusion is that if the British Early Iron Age is not to be explained by the arrivai of Hallstatt migrants it is equally unlikely to be understood in terms of the steady, indigenous growth of local agricultural economies. The Late Bronze Age must have been one in which the movement and resettlement of people and the re-definition of entire conununities was a consequence of the ways local agricultural economies were linked, and to a large extent dependent upon, complex networks of exchange relations. The implication is that change in the dominant metal technology may have had profound implications for the operation of these population dynamics.

Fig. 4. Plan of the area around the hillfort of Sidbury on Salisbury Plain. The earliest phase of the hillfort is contemporary with the final stages of the use of East Chisenbury midden and is associated with a number of linear boundaries. These linear earthworks slight earlier fields and are focussed on a range of smaller settlement sites as well as earlier features such as round barrows.

17As our understanding of the historical complexity of the period has developed, so too has the récognition of the range and the complexity of the surviving archaeological deposits. This has resulted in part from the realisation that archaeological deposits survive far more extensively than was once thought, due to the considerable increase in the scale of recent excavations, and the extension of excavations into previously under-explored régions. This recent work mainly results from greater levels of protection afforded to archaeological sites in the face of development and the conséquent investment of developer funding into archaeological field work. At the same time previously unrecognised types of archaeological deposit have also been identilied, among which are a sériés Early Iron Age « midden sites » . These sites are important because they may provide a context for an understanding of the earlier excavations at All Cannings Cross.

18Deep midden deposits belonging to ail periods of prehistory have long been known from the northern and western isles of Scotland where such accumulations within Coastal dune Systems can be accounted for by the protection provided by the blanketing of blown sand. What is surprising about the discovery of deep deposits in Southern Britain is that they have survived unrecognised in régions that hâve been intensively surveyed by archaeologists for several generations and which have also been the focus of intensive, and therefore destructive, agriculture. For most of the twentieth century the investigation of prehistoric settlement has prioritized the most visible sites, initially those surviving as earthworks and later as crop-marks revealed by aerial photography. This has meant that the former directed attention to marginal upland régions of high survival, the latter to the well drained river terraces where the archaeology is heavily plough eroded. Neither environment is conducive to the survival of occupation surface deposits, even where residues of such material might survive the archaeological use of mechanical excavators will not hâve helped the recovery of such traces. What we have tended to overlook are the less visible but nonetheless well preserved deposits buried beneath hill-wash or alluvium where occupation deposits are both masked and protected. Runnymede Bridge on the Thames embankment is one such example; here, well preserved Late Bronze Age occupation surfaces have been protected by a deep covering of alluvium (Needham and Spence 1996).

Fig- 5- View of the earthwork mound at East Chisenbury, viewed from the south. The mound is circular in outline and survives to a substantial height despite the impact of later cultivation as well as the impact of military training activities.

19Between 1982 and 1992 two substantial deposits ( « middens ») of humic material, animal bones, pottery and metalwork were investigated in the landscape around the Vale of Pewsey. One site at Potteme was first noticed when grave diggers, cutting through the then unrecognised archaeological deposit, recovered a Late Bronze Age gold bracelet prompting a programme of archaeological excavation (Lawson 2000). The second site of East Chisenbury was identified as part of an archaeological field survey although finds had been recovered from the area some forty years earlier (McOmish 1996). Neither site was deeply buried, indeed East Chisenbury has long been under cultivation although it still survives as a mound of material up to at least 2 métrés deep and covering almost 3 hectares in area (fig. 5). The Potterne deposit is also around 2 metres in depth and covers around 3.5 hectares. Both sites were unenclosed by earthworks and East Chisenbury had not been recognised simply because of its scale: at first sight the mound appears to be more a feature of the natural topography than a humanly constructed feature. While Late Bronze Age metalwork occurs on both sites the recovery of iron artefacts and the nature of the ceramics would centre the chronology of them in the earliest Iron Age. Both excavations were confronted with attempting to understand the accumulation of humic deposits, rich in artefact and animal residues but with a complex sequence of stratigraphy that is extremely difficult, if not impossible, to distinguish visually. In the case of East Chisenbury the one clearly observed feature of the stratigraphy was the existence of a sériés of chalk « floor » deposits and spreads of burnt clay laid within the midden build-up (fig. 6). Although both sites appear similar, interpretations differ. Lawson argues for a relatively long chronology for the Potterne accumulation across the Late Bronze Age/Iron Age transition, although it is also possible on the basis of finds and the radiocarbon dates, to suggest a shorter period of activity with the main deposit developing within century or so of activity. Lawson also regards the accumulation of material as resulting from local, every-day agricultural and domestic activity. On the other hand, East Chisenbury seems to hâve resulted from a rapid accumulation forming perhaps within a century, that included the deliberate dumping of bone, organic waste and artefacts (fig. 7).

Fig. 6. West section of Trench B during excavation at East Chisenbury. The chalk deposits, seen here and noted at All Cannings Cross and Stanton St. Bernard too, defy explanation but it may well be that they are related to former structures within the midden build-up – either as depositional platforms or as demolished buildings. In all cases they seal substantial deposits of pottery and bone.

Fig. 7. Diagnostic range of artefacts from the excavations at East Chisenbury. The assemblage is dominated by pottery forms and metalwork that dates closely to the 8th – 6th centuries calBC. The most distinctive components are the short- and long-necked furrowed bowls (illustrated by the two substantial ridged sherds in the centre of the photograph) as well as the large decorated coarseware jar (standing at the rear).

Fig. 8. Linear earthwork on Salisbury Plain. This significant earthwork boundary, which probably dates 10th/9th century calBC, extends for several kilometres across open downland and separates the lower-ground and river valleys to the south, from the higher chalk plateau to the north. It is aligned on a Neolithic long barrow (top, centre) and the major angle change at its midpoint ensures that it avoided a pre-existing settlement. In several locations along its length it overlies earlier field System.

ALL CANNINGS CROSS AND STANTON ST. BERNARD 2003-2004

20If the validity of the above arguments are accepted then the founding of the Southern British Iron Age is part of a complex set of processes involving rapid changes in settlement and landscape organisation and the re-orientation of long distance trade networks. It would be worth noting, therefore, that whilst very early occurrences of iron working may occur in some places, the full adoption of that technology will have required a number of wider social and économic factors to be in place.

21All Cannings Cross, Potterne and East Chisenbury are all situated around the Vale, and all lie within 18 kilometres of each other. These sites therefore provide a significant focus for our understanding of the Late Bronze Age and Iron Age transition at a regional level. With regard to changes in settlement and landscape organisation, there is certainly evidence on the Marlborough Downs, to the north of the Vale, of Middle Bronze Age landscapes that had been abandoned by the Late Bronze Age (Piggott 1942), and the linear ditch Systems of the Late Bronze Age also imply a reorganisation of the landscape on Salisbury Plain to the south (fig. 8). If the Late Bronze Age and Early Iron Age is a period in this region that witnessed not only a change in metal technology but an ongoing process of settlement reorganization (including the founding of the earliest hillforts) and changes in agricultural management, then are the recently identified midden accumulations indicative simply the accumulation of domestic debris or the resuit of the more specialised and major acts of consumption as suggested at East Chisenbury?

22All Cannings Cross now cornes back into focus with its potential to throw light on this new set of issues. Although the original report is difficult to interpret in any detail, certain general observations can be made. Cunnington records excavating through deep humic deposits, some of which were very rich in finds and the pottery recovered from her excavations includes much material surviving as large sherds, some comprising material from near complete vessels. The combination of humic deposits and this quality of artefact recovery is strikingly reminiscent of the character of material in the recently excavated midden sites. However, All Cannings Cross has also produced a wide range of structural features associated with settlement activity, including pits, post-built structures and laid surfaces or floors of chalk stone along with surfaces of burnt clay. Some of these chalk surfaces overlay humic deposits, recalling the chalk and clay spreads noticed in the East Chisenbury midden. The pits include examples eut into the slope of the hillside, others cut down through apparent occupation debris and chalk surfaces. The pits may have satisfied a range of functions; some were capped with clay and some appear to have been associated with firing processes, in one case accompanied by the deposition of a large amount of iron slag.

23The site thus appears to combine deep stratigraphy, complex settlement activity and deposits reminiscent of the accumulations of material recorded on neighbouring midden sites. There are however problems. We do not know how fully Cunnington’s work force excavated the area she records as being trenched, and we do not know the full extent to the site itself. There also remained doubts about the survival of these deposits today given that the site had remained under cultivation until relatively recently. In 2003 we therefore undertook reconnaissance excavations with the aim of assessing the level of survival and potential of any remaining deposits. This meant that we did not intend, at this stage, to completely excavate a sequence of deposits to a buried, « natural » land surface.

24The site lies at the foot of the chalk slope with the land rising steeply to the north and west. It is overlooked by the hillfort of Rybury, and it looks out over the Vale to the south and east (fig. 9). In this broad and deep basin of land it is quite clear, simply from the vegetation and surface soil colour, that while the chalk bedrock is both close to the surface and eroding on the slopes, the bottom of the basin is of quite a different character containing a very deep accumulation of soil. Our excavation technique was to dig a series of 1 metre square test pits, supplemented by augering, across and beyond the area covered by Cuimington’s excavations (fig. 10). In two areas these test pits were extended into larger areas of excavation. This work clearly revealed surviving archaeological deposits across the whole of the area investigated, although the survival of these deposits varies as the result of erosion and the earlier excavations, with evidence of deposits at the bottom of the slope sealed and protected by hillwash resulting in up to 2 metres of total accumulation. In one area we were able to partly excavate a laid chalk surface which had been placed over a spread of animal bone and pottery. However, in all the areas we investigated nowhere did we encounter the large unweatherd sherds of pottery recovered by Cunnington, indeed ail the pottery recovered in 2003 was small and eroded. The question remains: did Cunnington encounter a midden type deposit within a larger area of settlement? It would certainly appear from both excavations that different kinds of deposit occur across the site, perhaps with smaller and less substantial spreads of occupation debris alongside more substantial and undisturbed deposits of primary refuse.

Fig. 9. Location of the deposit at All Cannings Cross. It sits at the base of a prominent chalk escarpment and is overlooked by the prehistoric enclosure complex of Rybury.

Fig. 10. Students from Sheffield University digging a 1m² test pit on the eastern flank of the All Cannings Cross deposit.

Fig. 11. Test pitting at Stanton St Bernard.

Fig. 12. The Stanton St Bernard deposit differed markedly froin both Ail Cannings Cross and East Chisenbury. Analysis of the soil suggests that a large part of it consisted of burnt material and ashy soil resting on top of a well developed forest soil; there were indications that the deposit had been cultivated during its development. It is of similar date to the earliest phases at All Cannings Cross and East Chisenbury.

Fig. 13. Battlesbury hillfort on the south-western edge of Salisbury Plain dates to the middle centuries of the 1 st millennium BC. It sits on top of an extensive Late Bronze Age and Early Iron Age site that sharedmany characteristics of the deposits observed at East Chisenbury et al.

25During this first season’s work, reference to the available aerial photographic cover revealed that the high organic content of the soils at All Cannings Cross rendered the site visible, whilst under cultivation, as an extensive black soil mark. The same coverage revealed a similar soil mark, also situated against the foot of the chalk, at Stanton St. Bernard, 600 metres to the east of All Cannings Cross. Fieldwalking and a subsequent season of reconnaissance excavations, employing procedures similar to those used at All Cannings Cross have revealed this to be another midden of the eighth century BC (fig. 11). The mound is visible from the ground, and the deposits survive to a maximum depth of 1.5 metres and cover an area of at least 2 hectares. The mound is rich in animal bone and unabraded pottery sherds (fig. 12).

26We have argued that the eighth to seventh centuries in southem Britain can be characterised as the culmination of a wide range of changes in landscape, settlement, trade and craft organisation and that these provided the context for the full adoption of iron working. The complexity and dynamic of these historical processes are not adequately accommodated by the models of continuity that were developed in the shadow of Hodson’s critique of Hawkes’model for the British Iron Age, and they will have had geographically widespread implications for population dynamics and cultural identifies. Nonetheless, the particular patterns that are now emerging for the Vales of Pewsey should not be taken as more than the regionally specific example of these more general developments. The region, with its early hillforts along the northem downs, is likely to hâve witnessed a number of quite particular forces that drove a specific trajectory of development, one such being the availability of an easily worked and high quality iron source at Seend, 17 kilometres to the south-west.

27The extensive middens of this date at Potterne, Stanton St. Bernard and East Chisenbury appear, on present evidence, to be examples of a larger number of such sites around the Vale of Pewsey in central Wiltshire. Their implications for the agricultural productivity of the hinterland are considerable, representing as they do the slaughtering of many thousands of animals as well as the consumptions of other agricultural and craft products. One priority for future work must be to attempt a quantification of these levels of consumption and to consider the historical and social contexts in which this kind of activity is likely to have occurred. It also remains a priority to establish the nature of the settlements associated with these sites and to consider if the building of the early lillforts was accompanied by a dislocation of seulement activity within the wider landscape (fig. 13). It is therefore possible that All Cannings Cross, originally identified as a cultural assemblage to define the beginning of the Iron Age, might now have a rôle to play in establishing the form of seulement activity that accompanied technological change and the extensive exploitation of agricultural productivity.

Acknowledgements

28We gratefully acknowledge the support, interest and enthusiasm of Tim Daw at All Cannings Cross and Brian Reed at Stanton St. Bernard for permission to excavate on their land. The programme of work would not have been possible without the financial support provided by the Royal Archaeological Institute and the University of Sheffield, and the administrative support provided by Ms Cat Howarth. Both season’s fieldwork were supported by the advice of Dr Mike Allen of Bournemouth University and Dr David Norcott of Wessex Archaeology; their input is, again, gratefully acknowledged here.