Lives in Land – Mucking excavations: Volume 1. Prehistory, Context and Summary

By Christopher Evans, Grahame Appleby and Sam Lucy

The excavations led by Margaret and Tom Jones on the Thames gravel terraces at Mucking, Essex, undertaken between 1965 and 1978 are legendary. The largest area excavation ever undertaken in the British Isles, involving around 5000 participants, recorded around 44,000 archaeological features dating from the Beaker to Anglo-Saxon periods and recovered something in the region of 1.7 million finds of Mesolithic to post-medieval date. While various publications have emerged over the intervening years, the death of both directors, insufficient funding, many organizational complications and the sheer volume of material evidence have severely delayed full publication of this extraordinary palimpsest landscape.

Lives in Land is the first of two major volumes which bring together all the evidence from Mucking, presenting both the detail of many important structures and assemblages and a comprehensive synthesis of landscape development through the ages: settlement histories, changing land-use, death and burial, industry and craft activities. The long time-gap since completion of the excavations has allowed the authors the unprecedented opportunity to stand back from the density of site data and place the vast sum of Mucking evidence in the wider context of the archaeology of southern England throughout the major periods of occupation and activity.

Lives in Land begins with a thorough evaluation of the methods, philosophy and archival status of the Mucking project against the organizational and funding background of its time, and discusses its fascinating and complex history through a period of fundamental change in archaeological practice, legislation, finance, research priorities and theoretical paradigms in British Archaeology. Subsequent chapters deal with the prehistoric landscape, each focusing on the major themes that emerge by major period from analysis and synthesis of the data. The authors draw on archival material including site notebooks and personal accounts from key participants to provide a detailed but lively account of this iconic landscape investigation.

• Language: English
• Publisher: Oxbow Books
• Release date: Dec 31, 2015
• ISBN: 9781785701498

Christopher Evans

Having worked in British archaeology for over thirty-five years, Evans co-founded The Cambridge Archaeological Unit, together with Ian Hodder, in 1990. He has directed a wide variety of major fieldwork projects, both abroad (Nepal, China & Cape Verde) and in UK, most recently publishing the results of the Haddenham Project in 2006, the South Cambridge/Addenbrooke’s Environs (2008), Fengate Revisited (2010) and the Colne Fen Project’s Process and History volumes (2013). Elected a fellow of the Society of Antiquaries of London in 2000, he is a member of editorial board of The Bulletin of the History of Archaeology and, together with Tim Murray, edited Histories of Archaeology: A Reader in the History of Archaeology for Oxford University Press (2008).

Book preview

Chapter 1

Introduction

Landscape and Archival Palimpsests

Set against the background of A Matter of Time’s call to arms concerning the threat posed by gravel quarrying to the archaeology of Britain’s river valleys (RCHM 1960), everything started with St Joseph’s 1961 aerial photographs of the area (Fig. 1.1). Published in Antiquity in 1964 in relationship to the Dorchester Rings (together with Rudston, Yorkshire’s vaguely comparable ‘double-circle’), St Joseph interpreted Mucking’s concentric circular enclosure as a Neolithic henge. The paper concluded by noting that on-going gravel workings were visible in the image’s lower left corner and, given their progress since first flying the area in 1959, that the monument should either be saved or carefully investigated (St Joseph 1964, 218; see also Riley 1987, 96–8 and in Clark 1993, 23–5 concerning Mucking’s cropmarks).

Accordingly, the cropmark complex was brought to the notice of the Ancient Monuments Section of the Ministry of Public Building and Works with the aim of scheduling 40 acres. Although unrealised, this was sufficient to prompt the excavations that commenced in September 1965. Progressing northeastwards with little interruption, the fieldwork was to continue until 1977 and, eventually, extend over some 18ha (there also being a brief reprise in the following year to deal with the North Ring-area; see below). With upwards of 5000 participants involved over the years (many from overseas), this was rescue archaeology par excellence, and the staggering scale of Mucking’s excavations is told by its gross statistics. Aside from eight round barrows and a Bronze Age fieldsystem, more than 110 Late Bronze/Iron Age roundhouses were recovered and, interred within formal cemeteries associated with settlement compounds, were some 170 burials of the Romano-British period. Perhaps most important was the scale of its Anglo-Saxon occupation; accompanied by at least 57 post-built ‘halls’, more than 200 sunken-feature buildings or Grubenhäuser were excavated, as were also some 800 contemporary burials.¹

With the double-circle enclosure – the South Rings – proving quickly not to be of Neolithic date, but rather of later prehistoric attribution (initially thought a ‘mini’ Iron Age hillfort and only later shown to be Late Bronze Age), Mucking’s Anglo-Saxon archaeology quickly dominated the project’s agenda. Related by its excavator to Continental folk migrations after AD 400, the latter phase came to determine the site’s setting and from the early years of the project its location maps were shown in relationship to the adjacent North Sea Coast countries (Fig. 1.10.3; Jones et al. 1968, fig. 1).

Of its more immediate setting, as shown on Figure 1.2 the site extended along the Boyn Hill Terrace between 29–34m OD. Along its eastern margin, the ground drops steeply (where the terrace’s gravels are covered by brickearth) down to the marshes and flats alongside the Thames (see Fig. 4.87 and Carreck’s outline of the area’s geology in Clark 1993, 26–31, fig. 3; see also Murphy, Chap. 2 and Rippon, Chap. 6, below). Its strategic situation – near a low natural crossing point of the Thames, and at the point where it turns markedly southwards and begins to narrow from its broad estuary mouth – was crucial; literally, at the start of the river per se and the ‘gateway’ into southern Britain:

This is a place which would naturally suggest itself as landfall to travellers up river, whether traders, immigrants, or invaders, whether from south-east coastlands or from the Continent (Jones et al. 1968, 212).

Yet, along this stretch of the Thames many locations fulfil these broad criteria and where exactly ‘the gateway’ – like the river (vs. its estuary) – begins and ends is not absolute. To be frank, set c. 3km back behind the river proper would not seem an ideal strategic locale; for settlement perched on a 30m-high escarpment immediately above marshland and estuarine flats, access to wetland resources may have been as much an attraction as ‘strategic’ concerns. That said, as will be explored in this book’s final chapter, with its view-sheds along the river and the potential for observing arrivals from the Continent, Mucking’s location was unique. Indeed, a theme of our final chapter is that the dynamic sweep and historicist character of the site’s original interpretative framework has now – after intervening decades of paradigms focused on insularity and continuity – in some respects, again proven relevant in the light of its Thames Gateway locale and the demonstration of past cross-channel connections.

From the outset it should be stressed that in covering Mucking’s background, excavation/recording techniques and post-excavation history, portions of what appear in this chapter are a ‘twice-told tale’ and have been summarised before (see Clark 1993 and Hirst & Clark 2009). Yet, as befitting books appearing in a series entitled Historiography and Fieldwork (see below), in many respects our concerns and ‘take’ differ considerably. This is especially true of our focus on the site’s archival sources – in other words, ‘land-into-text’ – and the nature of archives generally and historically contingent interpretations. Issues of excavation precedent and expectations are crucial for understanding Mucking, and these run almost as a refrain within this volume. A ‘comparative technique’, involving long-accrued parallels, sits at the heart of almost all archaeological practice (see e.g. Evans 2012 further to this theme). Although now readily taken for granted, in order to understand Mucking’s approaches and explanatory framework, it has to be appreciated what it was to do fieldwork at a time when there was so little direct comparative context and the challenges that entailed (i.e. so few other sites had then been exposed to anything like a similar scale). To do the Joneses any justice, these matters need to borne in mind throughout.

Total Archaeology

Often harking back to Mucking’s discovery through aerial photography, while extolling the virtues of that technique, at the same time Margaret repeatedly emphasised the need to actually dig the sites/landscapes that it revealed. She stressed the surprises that excavation alone threw up and was obviously wary that a ‘school’ of cropmark-enclosure morphology could over-ride the need to actually dig them (e.g. Jones 1972a).

The original aim was for 100% excavation of Mucking’s fill-deposits, though, in practice c. 75% was generally achieved (Figs. 1.3–5; see Fig. 4.5 and Clark 1993, fig. 5). While by today’s low-density standards this seems extraordinary, it was, for example, achieved at Dragonby (Fig. 1.7; May 1996), and even Fengate in the 1970s was very intensively excavated (Pryor 1980a). There was then, obviously, little notion of formalised site sampling procedures and, rather, maximising finds retrieval appears to have been the guiding factor: a logical goal at a time when many regional artefact typologies were still being established. ‘Rarity value’ may well have also been influential. Despite the fact that aerial photography was beginning to suggest, at least locally (i.e. on gravels), how extensive sites were, there was very little concept of the actual density of pre-Medieval settlements across much of southern Britain. With far fewer sites known (and expected), the loss of any – and even portions thereof – would clearly then have been held to be a far greater loss than today. Nonetheless, the greater Mucking landscape had obviously seen considerable post-War destruction as wrought by industrial and housing estates, as well as brickearth pits and gravel quarries, with little or no archaeological recording; the site, thereby, presented an opportunity to rectify matters.

Mucking saw one of the first applications of the term ‘landscape palimpsest’ in British archaeology (with Crawford 1953, 51 seemingly marking its first usage; see Bowden 2001, 42–3). Occurring in a 1968 Panorama interim (i.e. the Journal of Thurrock Local History Society) and, thereafter, featuring in the title of a paper five years later (Jones 1973), this implied that its landscape features were the accumulation of multiple phase-usage and not the relic of a single horizon. Certainly, the usage of this term was without any manner of ‘knowing’ textual metaphor-reference (i.e. ‘over-writing land’), ‘erasure’ or ‘time perspectivism’ (e.g. Bailey 2007). What constituted the ‘landscape’ portion of this description is rather what counted – a sense of sites/monuments without borders:

It would be well to consider the nature of such an extensive ‘ancient monument’ as the forty acres of the Mucking crop mark sites. In the first place, it is an area and not a finite building site or earthwork, such as a monastery or a Bronze Age barrow. The knowledge of its limits is defined only by the extent of the crop marks themselves. This is clear when one traces the crop marks of boundary ditches. Both west and east along the gravel terrace, and also down its slope, ditches continue beyond the photograph. In the same way that such a monument has no clear boundaries in space, so does it lack precision in time. Crop marks can develop over a modern sheep burial in the same way as they do over a Bronze Age pit. These crop marks show then not just one landscape, but a palimpsest of landscapes (Jones 1968, 36; emphasis added).

As, in effect, the project’s main conceptual/interpretative background, this should caution us against over-intellectualising Mucking. It was framed by the abiding rescue archaeology ethos of the 1960–70s, but actually anticipated the foundation of RESCUE by seven years (of which the Joneses were founding members).² As opposed to, for example, David Clarke’s use of the phrase ‘Total Archaeology’ in relation to his highly ambitious Great Wilbraham investigations (also having over-arching aims resting on minimal infrastructure; see Evans et al. 2006), in the case of Mucking it was simply considered the appropriate response to the scale of the site’s destruction: ‘total destruction warranting total excavation’.

Figure 1.1. Starting Points: St Joseph’s 1961 aerial photograph (top) with A Matter of Time cover imposed; left, Mucking Section 1 (South Rings; note the indication of differential crop growth over ditch in demonstration of cropmark register).

Figure 1.2. Landscape profile (with exaggerated vertical scale; below) and location map: 1) Mucking excavations; 2) North Ring; 3) Linford; 4) Rainbow Wood; 5) Orsett ‘Cock’; 6) Orsett causewayed enclosure; 7) Rectory Road, Orsett; 8) Stanford Wharf/London Gateway Compensation Site A (after Clark 1993, fig. 2, with additions).

Otherwise, and not surprisingly for the time, Mucking was essentially interpreted within a cultural-historical paradigm, with long-distance trade, ‘folk movements’ and, for its Iron Age, transhumant shepherds, providing explanatory mechanisms. Perhaps reflective of her background in geography (see below), from the project’s early days Margaret was much given to large-scale mapping, plotting the wider spread of the site’s find-types (e.g. Jones 1972a), and on more than one occasion cited the importance of Fox’s distributional studies (Fox 1923). As has been discussed elsewhere (e.g. Evans 1987), such dynamics were, of course, also logical at a time when there had been relatively little excavation: without recourse to movement, how else would you join up far-flung distribution dots?

In the case of Mucking this emphasis had a greater resonance due to its ‘gateway/frontier’ situation, especially in the light of the scale of its Early Anglo-Saxon presence. Yet, this approach to the site’s sequence coincided with a sea-change in archaeological interpretation, with the publication of Grahame Clark’s renowned ‘anti-invasionist’ paper occurring in 1966 (see Chap. 4 further to this theme in relationship to the project’s ‘cognitive map’; Fig. 4.2). The same era also saw the early impact of Processual/New Archaeology. Though David Clarke’s Beaker studies were cited in relationship to Mucking’s burials of the period, early papers by, for example, Binford – calling for more explicit methodological direction and problem-orientation (e.g. 1968) – were not.³ This, of course, is not at all surprising and such approaches really had no impact in Britain until the latter half of the 1970s (e.g. Cherry et al. 1978). Margaret was, nevertheless, to prove quite intractable in her ‘anti-theoretical’ stance. As outlined in the recent Fengate Revisited volume (Evans et al. 2009, 250–2), she took extreme umbrage at Pryor’s overtly problem-based and methodologically driven approach to excavating Maxey (Pryor 1980b).

Margaret was clearly more concerned with matters of ‘dirt archaeology’ – questions of excavation/recording methodology and site formation processes – as well being fully committed to the popularisation of fieldwork. Aside from the regular issue of local society interims and notes (even throughout post-excavation), much newspaper and television coverage, public lectures and local displays, the latter had expression in the site’s postcard series (Fig. 1.16 showing just one of the four that were issued).

The Joneses were certainly not, however, insular in their approach to Mucking. Although like many of their generation and background they were not well travelled, in 1961 they had apparently accompanied Higgs to survey Greek Palaeolithic sites (Elsdon 1997, 78). They had professional ties with Continental scholars, touring Denmark and Holland in search for the homeland links of Mucking’s Saxons (ibid. 79), and Margaret even published a summary of the site’s pottery of that date in a Dutch journal (Jones 1969; see also 1978c).⁴

Mucking’s archives also reflect a commitment to the burgeoning archaeological science of the day. This arose through close working with the Ancient Monuments Laboratory and also the opportunistic courting of specialists. The site, for example, saw early resistivity trials (Fig. 1.7) and there was offsite survey to retrieve clay samples to test against its briquetage (see Barford & Mainman, Chap. 3). Similarly, pottery was submitted for neutron activation analysis and correspondence files show that Margaret had contacted Dutch colleagues who had worked with pottery diatoms to establish whether their clays were saline (02/08/79; see also Jones in Clark 1993, 9). The excavations also included an experimental component. Pots were made by students from the local brickearth clays (and were apparently test-fired in kilns) and, in relationship to what might be considered the overly wide span of roundhouse roofs, there is a citation that this was being addressed through ‘practical experiment’ (Jones & Jones 1975, 141–2).⁵

In contrast, the site’s economic data was not greatly prioritised. This is not at all surprising, nor uncommon at the time; after all, the project’s inception predated the call to arms of the palaeo-economy ‘school’ by some seven years (Higgs 1972). Mucking’s bone preservation was certainly very poor. This was true of both its human burials, which usually only survived as ironpanned casts/stains, and also its animal remains; of the latter, only some 10,000 pieces were apparently recovered. Although a great number of soil samples were collected (934), it would seem that the carbonised grain from only 25 was ever submitted to the Ancient Monuments Laboratory for analysis and that little flotation was actually attempted. This was not at all unusual for the time and it is clear that they were, instead, essentially relying upon in situ/’macro-observation’ of burnt plant remains and, otherwise, their impressions within fired clay and pottery (see van der Veen in Hamerow 1993, 80–1); specific finds categories would also provide a basis to infer economic activities (e.g. spindlewhorls and loomweights). Against all this, it seems extraordinary that, early in post-excavation, they had the site’s many thousands of charcoal fragments identified and their distributions were among the first of its computer plots. (This was due to the immediate availability of an appropriate specialist; i.e. Grahame Morgan, see Chap. 3.)

Figure 1.3. Site plan, 1980 (from Clark 1993, fig. 4).

Figure 1.4. Digital site plan and contour map.

Figure 1.5. Total Archaeology: The Working Face-edge: left, looking down the ‘Double Ditch Enclosure’ line and, below, the ‘Belgic Banjo’ Area (see also Fig. 4.58).

Figure 1.6. The Missing: With its area indicated in red on the 1973 interim plan (upper right), the 1969 aerial photograph (left; RAF V58/9640 004) shows an east–west swathe stripped across the central site (with the southern Belgic Banjo compounds top and the near-proverbial excavation campsite lower right; see Fig. 2.44) and is disturbing on a number of counts. First, with the quarry’s drag-line progressing along the left/southern side – with almost nothing in front of it as yet excavated – it conveys just what it was they were up against and the relentless pace of the work (especially given the limited size of the campsite/site staff when compared to what needed to be dug). Secondly, the photograph also has the qualities of surveillance, as it allows comparison between what was in the ground and what was actually recorded and dug. Unfortunately, in reference to the Atlas’ Map Sheets 6, 8 & 10, it displays considerable discrepancies and these should inform any appraisal of the site’s archaeological recovery. Among the ‘missing’ (i.e. unavailable for examination) are the circle of a roundhouse (1), a small sub-rectangular ditch-setting (possibly marking a Late Iron Age/Conquest Period building; 2), a curvilinear ditch relating to the Belgic Banjo’s sequence (3) and, at ‘4’ beside Roundhouse 51, possible further gully-set granaries (cf. RHs 56 & 58; see also Fig. 4.64); Mucking’s ‘missing’ were, though, accurately expressed in the 1973 plan’s blank-spaces and are further demonstrated in the site base-plan (lower right), with grey-tone indicating swathes of limited/non-feature recovery.

Figure 1.7. Dragonby, Lincs., Site 1 (May 1996, fig. 5.1) and, below, Mucking’s ‘science’: resistivity survey plot (1967; left) and, right, histogram of Grubenhäuser (Jones 1979). Another ‘total excavation’, the 1964–73 Dragonby, Iron Age and Romano-British settlement excavation (c. 0.7ha in total, of which Site 1 was its largest exposure; see May 1996, fig. 1.3) involved up to 80–100 volunteer staff per-season and cumulatively took 81 weeks (and cost, in total, £28,685, plus £170,000 for its post-excavation; ibid., 7). Also, like Mucking, seeing the application of much early-day archaeological science, as indicated on the plan it was also a ‘complete’ 100%-sample excavation. As such, it provides a further measure of total artefact-recovery rates: 14 human burials, eight Iron Age roundhouses, 10 Romano-British buildings, 157 brooches (89 Iron Age/Conquest Period), 249 coins (including 15 British), c. 150,000 animal bones and ‘about two tons’ of predominantly Iron Age and Early Romano-British pottery (if assuming a mean sherd weight of 20g, this would equate to c. 90,000 sherds). In many of its finds categories (as well as by area-density) its numbers were actually higher than Mucking’s, but as far as can be established Dragonby’s spatial control of finds by feature was not as precise as Mucking’s.

In her penultimate Panorama note, ‘Cereals at Mucking’, Margaret overviewed the site’s plant-use (1985a). Aside from discussing the pottery’s grain impressions, the sheer scale of the site’s quernstone assemblage, plus its raised Romano-British granary and corndriers, she also reviewed its four- and six-poster settings as Iron Age granaries (see, though, also Jones 1974, 190–1). In it she cited the evidence from recent Essex sites and, also, Gent’s review on that theme in the Proceedings of the Prehistoric Society (1983). If not for the demise of her direct post-excavation involvement (see below), whether this recognition would have been sufficient to alter her earlier, Iron Age ‘transhumant shepherds’ argument cannot be known. It does, however, perhaps suggest something of a redirection of her cultural-historical approach, perhaps brought about by a greater awareness of the potential of a broader range of the site’s data:

Although archaeology seems obsessed with finds and where they come from, the aim is essentially to shed light on how people lived in the past. So it is unfortunate that the Mucking post–excavation project has to close before all the finds, facts and figures, so laboriously amassed in the 1965 to 78 excavation, can be studied in their many themes and aspects (1985a, 17; emphasis added).

On the whole, the kind of interpretative ‘hotchpotch’ that Mucking’s sources attest to is arguably reflective of the general situation of fieldwork. On the one hand, sites are essentially a product of the milieu in which they are first conceived. In its scale, Mucking certainly showed far greater vision than most; the primary problem being that, over-extend a project’s lifetime, and its approaches will invariably become outmoded. At the same time, there is also clearly the influence of personal contacts and seemingly random source-exposure, plus the changing whims of central state finance (the latter being particularly exasperated by project ‘over-duration’). As further outlined below, for Mucking the impact of the 1975 Frere Report had the greatest consequences and, leading from it, the recognition of the possibilities of intra-site distributional analysis through the site’s computing during the later 1970s.

Framing Context

It is necessary to establish a sense of the project’s ‘framing context’. Before exploring the era’s broader excavation background, we need first to outline that of the Joneses themselves and, for this, we are fortunate to be able to draw upon Shelia Elsdon’s appreciation of Margaret and Tom which appeared in her Sleaford volume (Elsdon 1997, 77–80). Studying geography at Liverpool University, it was through W.J. Varley, who taught there and was an amateur archaeologist, that Margaret was introduced to digging and went on to volunteer at a number of sites (where she meet Tom Jones, a professional photographer and amateur fieldworker). Supporting herself though a variety of jobs – including roles as diverse as a Land Army administrator and cookery correspondent for the Birmingham Mail – it was apparently through participation on the O’Neil’s annual excavations on the Scilly Isles (Beagrie 1989) that, from 1947 onwards, Margaret decided upon a career in archaeology. In this she was encouraged and later sponsored by Helen O’Neil, an employee of the Ministry of Public Building and Works (and excavator of the Park St, St Albans Roman villa site; 1945), and her husband, Bryan, who was Chief Inspector of Ancient Monuments in the 1950s (Wainwright 2000, 909 and Rahtz 2001b, 604). On behalf of the Ancient Monuments Branch, in 1957 and 1958 Margaret directed work at the Iron Age and Romano-British settlement at Stanton Low, a gravel pit beside the River Great Ouse in Buckinghamshire. While immediately publishing the results of the initial 1957 trenching season (Jones 1958), the much more extensive excavations of the following year only saw publication 31 years later (and then, not by herself: Woodfield & Johnson 1989). Similarly, though publishing to high standard the 1964 trench investigations at the Roman fort at Aldborough, West Riding (Jones 1971b), aside from brief interim notices, it was left to later researchers to issue the results of her main sites: Old Sleaford, Lincs. (1960–3; Elsdon 1997) and Roughground Farm, Lechdale (1957–65; Allen et al. 1993 and see also Jones 1978a and Darvill et al. 1986). This, in itself, is not particularly unusual. There was then generally minimal, if any, resourcing available for post-excavation and many practioners of that time accrued vast backlogs (see Butcher & Garwood 1994). In terms of Mucking, it does seem remarkable that she would embark upon such an ambitious enterprise with only limited post-excavation experience, but then everyone has to start somewhere (by the same token, Fengate was Pryor’s first sole-directorial responsibility).

At Old Sleaford she had undertaken excavations across five main areas and, including further trenching, this amounted to c. 3150sqm (see Elsdon 1997, fig. 5). In the first season there, as was true of the main Stanton Low investigations (Woodfield & Johnson 1989, fig. 3), the site was dug using a Wheeler ‘box-technique’; only in 1963–4 were larger area-exposures employed at Sleaford, first taking off the topsoil by machine (Elsdon 1997, 16). In respects, Margaret’s 1957–65 Roughground Farm excavations were comparable. Progressing from a watching brief to a trenched wall-chasing exercise, the latter were thereafter expanded to provide localised area-exposure of its spectacular Romano-British villa, with topsoil being mechanically stripped by the quarry company (Jones in Allen et al. 1993, 1–3, figs. 34, 36 & 37). Certainly, Margaret was held to be a very competent archaeologist and fieldwork director, and the quality of the reports later issued based on her archives attest to the standard of the recording systems she maintained.

Allied investigations, that variously lay ‘behind’, bracketed and were concurrent with Mucking, also need to be outlined. First, occurring immediately south of the main site, were the 1955 Linford Quarry excavations (Figs. 1.2.3 & 6.15; Barton 1962). Supported by the Public Ministry of Building and Works Ancient Monuments Branch, this was prompted by earlier findings of both quantities of Iron Age pottery and what appears to have been Roman kilns in the vicinity. The excavation techniques employed in the 1955 site are discussed below. Of its findings, aside from a series of Romano-British ditched enclosures and Early Iron Age pit/posthole clusters, a series of Anglo-Saxon Grubenhäuser (accompanied by contemporary pits and post-built structures) were recovered.

The second associated excavation, located c. 250m north of the main Mucking site, was the 1978–9 excavation of the later Bronze Age ‘North Ring’ (Fig. 1.2.2; Jones & Bond 1980 and Bond 1988). This involved controlled open-area excavation by the Central Unit over c. 2185sqm of the circle’s eastern two-thirds and, also, conjoining salvage excavations undertaken by Margaret and Jonathon Catton (11,565sqm: Fig. 1.9). Across the latter, in addition to further prehistoric features, were even more Anglo-Saxon Grubenhäuser (see Jones in Bond 1988, 45–51). Aside from indicating that Mucking’s ‘palimpsest’ extended much further along the Boyn Hill Terrace, these excavations are relevant in the evaluation of the respective merits of their diverse techniques. In the case of the North Ring’s investigations, a high density of postholes were recovered within the ‘controlled’ area, with few such minor features surviving within the box-scraped salvage zone (Jones in Bond 1988, 46, fig. 45); as outlined below, differential machining depths also clearly affected Mucking’s results.

Although its methodology was not detailed in the report, as far as can be established the earlier Linford investigations involved a rather sophisticated adaptation of the ‘Wheeler-esque’ box-excavation techniques to serve as a site-wide sampling grid (Fig. 1.8; Barton 1962, 57). Occurring over 165 × 335ft (c. 1.7ha), 4 × 4ft squares were hand-dug arranged along transects (consistently across the eastern three-quarters, with its western cover being more judgementally based; i.e. ad hoc ‘ditch-chasing’). Based on the presence of features, a number of these were then variously expanded into trenches and larger open-area exposures. Finally, during the course of quarry-machined reduction, features were also recorded across the intervening unexcavated portions. The site, thereby, demonstrates the virtues of both hand-dug and machining techniques, and whereas Anglo-Saxon posthole structures were only recovered within the former, the Iron Age features were largely only found through the machine-stripping. Of the latter, Drury in fact cited Linford’s findings in demonstrating that the region’s Early Iron Age roundhouses were post-built (and ‘un-ditched’) and then employed its results to identify comparable settings on Mucking’s plan (1978a, 45–6, fig. 2).

The last ‘allied’ site to take account of is the 1969–70 excavations at Gun Hill, Tilbury (Drury & Rodwell 1973). While lying some 2km southwest of Mucking, the interrelationship between the two is not only based on the character of its sequence and fieldwork conditions – extreme quarry-rescue of a major Middle/later Iron Age compound (also with Romano-British kilns and a Saxon Grubenhaus; Fig. 1.9) – which offers obvious parallels with Mucking, but that the work was undertaken by Paul Drury and Warwick Rodwell, both of whom worked with the Joneses (see, also, e.g. Potter 1974 as another Mucking ‘off-spring’, and Rodwell 1996 concerning the post-War history of the County’s archaeology generally).

Also of immediate relevance was a spate of excavations during the 1970s within the neighbouring Orsett environs: its causewayed enclosure and subsequent settlements/burials (1975; Hedges & Buckley 1978), the Orsett ‘Cock’ enclosure in 1976 (Carter 1998) and, in 1979/80, the many sites investigated along the line of the Grays By-pass (Wilkinson 1988). Beyond these, one other, much more recent, adjacent site also deserves specific notice, that being Oxford Archaeology’s 2008–9 Stanford Wharf investigations (London Gateway Compensation Site A; Biddulph et al. 2012). Extending over 44ha along the north side of Mucking Creek and the Thames mudflats (including one of Essex’s ‘Red Hills’; Fig. 1.2.8), its significance here is its complementary location down on the riverside marshes proper, off the terrace. Although excavations per se were limited to approximately a third of the area, in addition to watching brief monitoring throughout, they were accompanied by a programme of geoarchaeological sampling. For our purposes, the main result, which will feature in both this and the Roman volume, was evidence of Late Iron Age to Late Roman salt production (see, also, e.g. Drury 1973, 118–22 and Rodwell & Rodwell 1973 for earlier findings from the Stanford-le-Hope area).

Table 1.1. Stripped excavation-area exposures (1964–83).

As detailed in Chapter Six below, understanding of the site’s landscape context has greatly benefited from the recent publication of the National Mapping Programme’s aerial photographic results from Essex, which includes a Mucking area case-study (Ingle & Saunders 2011). Equally, and also further discussed in that chapter, have been a series of major Thames Bank-side projects in Kent, as well as the entire coastal/foreshore dimension that, post-Mucking, has been introduced into the region’s archaeology.

Of Mucking’s broader contemporary excavation background, Pryor’s Fengate (1974–8) would have to feature as further demonstrating the application of ‘mass’ open-area machine-stripping. Also, as regards changes in fieldwork practices, are the development of single-context recording and matrix-analysis techniques. Arising in response to the challenges of deep urban sequences – both at Winchester and then, after experiments at Carthage, Cirencester and London (see e.g. Harris 1979, Roskams 2001 and Collis 2010) – the threat to Britain’s urban archaeology was another, if not the prime, rallying cause of the Rescue movement. There were, nevertheless, a number of earlier attempts to move beyond notebook recording, of which Dragonby provides an example (May 1996, 20–1; see also Pryor 1984, app. 1 and Evans et al. 2009, 12–13 on Fengate’s early attempts at computing).

The later 1960s and 1970s were marked by a series of milestone projects within the region. As, however, shown on Table 1.1, aside from Fengate, none were of a scale to rival Mucking, and this remains true even if the scope is taken to a national level. In the site’s many interims, Margaret clearly promoted the notion that Mucking should be established as the ‘type-site’ for such large multi-period landscape exposures. Yet, aside from North Shoebury, Highstead (nr. Chislet, Kent), Fengate and a handful of others (Wymer & Brown 1995; Bennett et al. 2007; e.g. Pryor 1980a and 1984), few projects at the time followed the Joneses’ lead. This mode of working did not become widespread until the 1990s, when c. 15ha+ sites began to be regularly tackled, with Heathrow Terminal 5 being the prime exemplar (Framework Archaeology 2006). With more readily available machining, most of the landmark excavations of the later 1960s/1970s were, instead, specific-phase cemetery-/settlement-focused and, in effect, they can be understood as starting to build the ‘pattern book’ of site-types. To be set against this, though, it needs to be recognised that great swathes of archaeological landscapes must then have been lost to quarrying and major construction works.

In terms of the archaeological infrastructure of the time, it is salient that even a ‘textbook’ site like Durrington Walls was done on a fairly shoestring basis and that its excavation involved large numbers of volunteers (Wainwright & Longworth 1971 and Wainwright 2000, 913). Similarly, undertaken 12 years later, Fison Way, Thetford (4.2ha) was dug using a youth unemployment scheme (Gregory 1991). This was also true of the 1975–7 Highstead Quarry work – a consciously ‘Mucking-type’ site – which was dug on very meagre resources (and only published 30 years afterward; Bennett et al. 2007, xxi). Indeed, apart from Fengate, funded by Canada’s Royal Ontario Museum, while other candidates might be advocated, the first really ‘formal’ (quasi-) professionally organised rural landscape project in Britain would have to be the M3 motorway fieldwork of 1972–83. Though not involving any single large exposure akin to Mucking’s (i.e. a series of single sites, each of which were the approximate size of Chalton Down or Gussage All Saints: e.g. Winnall Down and Micheldever Wood; Fasham & Winney 1991), collectively it was a programme on a vast scale and one propelled by a formal project design (Biddle & Emery 1973; see also Fowler 1979 for early motorway archaeology generally).

With Grimes’ 1944 Heathrow excavations marking the earliest known example (c. 9740sqm; Grimes & Close-Brooks 1993, 308–9), the impact of the use of earthmoving machinery in archaeology, in terms of the scale of sites which could then be tackled, should not be underestimated (see Evans 1989b on earlier, non-machined open-area techniques; see also Richardson 1951 and Rahtz & Greenfield 1977 for other early machining applications). Yet, as with the application of any new technique, it was something that had to be learnt and experimented with to achieve ‘the formula’ (with availability of the right machinery being a governing factor). As a result, ‘how to’ manuals were issued (Pryor 1974). In general, Mucking’s machining was provided by the quarry company, Hoveringham Gravels Ltd, with its overlying deposits being removed by box-scraper. For part of the site, however – presumably to maintain annual excavation-funding⁶ – Margaret Jones persuaded the Royal Engineers to strip it (and also had them contour-survey the site-area; Figs. 1.4 & 1.13).

Figure 1.8. Allied Investigations (I): The 1955 Linford Site (after Barton 1962).

In all this wing-and-a-prayer contingency, they did not always get things right; but then, whoever does? Reproduced in Figure 1.6, the interim plan of Mucking’s exposure up to 1973 is significant for how it shows the site’s marked ‘absences’: where its dense peppering of minor discrete features were not recorded.⁷ The largely blank swathes variously reflect quarry haul-roads, where the stripping either went too deep and/or where there was only time for cursory salvage recording between stripping and gravel extraction. These limited/non-recovery portions are shown on a site-wide basis in Figure 1.6 (lower right), and they extend to over 4ha.⁸ Indicating the time it took to adapt to the quarry’s working methods over the southern pre-1974 part, these amount to at least a third of the overall area. While obviously severely compromising claims to ‘total archaeology/recovery’ and hampering interpretation across those portions of the site, this is perfectly understandable given the fieldwork’s circumstances. Here, it is not so much a matter of bemoaning such loss as stressing the sheer density of Mucking’s features. Certainly unparalleled at, for example, Fengate (aside from the limited area of the Cat’s Water sub-site; Pryor 1984), even if taking into account the great landscape projects of the last 10–15 years, no other ‘scaled’ excavation has had feature densities approaching these (though West Heslerton’s 7.5ha Site 1 of 1977–84 came close; see Powlesland 2011). It is these densities of pits and postholes, as much as anything, that must sit at the core of Mucking’s long-term settlement sequence. As opposed to being able to readily characterise and roughly phase the site’s more major cut features (e.g. wells, Grubenhäuser and kilns) and various ditched configurations (occupation compounds/’folds’ and variously gullied roundhouses; see below), their attribution and articulation with the more obvious settlement settings has remained a crucial problem throughout. This is as true of the need to wed the site’s Anglo-Saxon Grubenäuser with its ‘halls’ and pits as it is for the Iron Age roundhouses and four-posters (and pits). Indeed, this is even apparent in the 1978 North Ring’s plan (Fig. 1.9), where only a minor portion of its array of pits and postholes were in anyway ‘interpretable’. As is strikingly portrayed on Figure 1.19 (its thousands of unassigned postholes being further discussed below), this was evidently something that the Joneses were fully aware of and in the 1974 interim report, after enumerating the structures and postholes then recovered to date, Margaret Jones noted:

The small number of structures plotted from the posthole scatters requires comment. Their total is deliberately conservative, first, because of the problems of isolating and dating posthole settings in a multi-period settlement, where the holes may contain no finds, indeterminate or residual finds, or undistinguished fills, and secondly, because of the reluctance to claim groups as buildings which may not be substantiated. It is not difficult to see pairs or even lines of three postholes in fairly dense scatters; they may, however, be more of a statistical hazard than of archaeological significance. But when excavation is completed, and the settlement foci and the orientation of different periods have been clarified, it is expected that many more postholes will fall into structural place. One could then legitimately indulge in what van Es so candidly calls ‘granary making’ (Jones 1974, 187).

Figure 1.9. Allied Investigations (II): Upper left, Gun Hill, 1969–70 (Drury & Rodwell 1973, fig. 2); upper right and below, Mucking, North Ring, 1978–9 (Bond 1988, figs. 3 & 30).

Figure 1.10. Scrapbook Sources: set against the background of the site’s 1966 Times coverage (left; with 1973 and 1976 digteam portraits) and one of the 1993 Atlas plan-sheets (right, also with ‘Notes for Excavators’ imposed) are Margaret (1) and Tom Jones (2) in action; 3) Mucking-to-North-Sea-Coast location map (Jones 1968, fig. 5); 4) excavation of the Iron Age square barrows (see Chap. 4, ‘The Plaza’).

The citation to the work of van Es at Wijster (1967) is salient, as a number of large-scale settlements had previously been dug on the Continent, with Feddersen Wierde or Köln-Lindenthal among the more obvious. These, however, were largely single-period exposures and lacked Mucking’s dense multiple-phase complexity.

Before progressing, the North Ring’s base-plan raises further issues and these will surface on more than one occasion in this volume. First, is the inherent ambiguity of its ‘backgrounded’ features. While a number of the site’s identified structures appear rather suspect (e.g. [1389]/[1730] and [1842]), others still also seem apparent (a sub-rectangular setting in the northeast corner, with further roundhouses and a four-poster suggested by the posthole array adjacent to [1708] and [1850]). Second, is the manner in which the site has featured in recent spatial/social structure studies. In both Brück’s 2006 ‘Fragmentation’ paper and Parker Pearson and Richards’ earlier ‘Architecture and Order’ contribution (1994, 50–1, fig. 2.6), it is the simple formality of the ringwork’s plan that has attracted interpretative speculation (i.e. its aligned roundhouses and ‘screen’), with its ‘extraneous’ features largely ignored. This is to the point that the later paper only really highlights the ringwork’s entrance-adjacent cremation – omitting the enclosure’s other two such interments (Brück 2006, fig. 4) – as it alone readily speaks. Sadly, it is almost as if academia now does not want too much site data. This is an ethos that can make one despair of the efforts it takes to disentangle an excavation’s ‘small feature’ detail and even raises questions concerning the effort that archive generation/maintenance entails, as they seem to be so rarely consulted in favour of published sources (see e.g. Dixon 1986 and Jones et al. 2001).

Notebook Archaeology

If the world and research were logical it could be argued that the advent of landscape archaeology in Britain would have had no better starting point than digging, analysing and publishing such a dense ‘palimpsest’ as Mucking in its entirety. However much the expected norm of the day, given the project’s limited resource base the attempt to do so shows remarkable audacity on the part of the Joneses. Yet, its burials, kilns and Grubenhäuser aside, the argument could be mounted that Mucking did not actually warrant such intense excavation due to its poor economic data – especially its bone preservation. This, of course, is an appraisal that can only be made with the benefit of hindsight and there would have been insufficient excavated precedent at the time to make such a judgement. It must, moreover, be appreciated that Mucking was a project that grew and, in trying to keep ahead of the damage wrought by the quarry, its scale was never envisaged (or planned for) from the outset.

The intensity to which the site was dug can be appreciated in Figure 1.5’s photographs (remembering that the intervening baulks were also subsequently excavated). A pressing concern with such ‘totality’ is the risk of pixilation. Arguably a case of ‘more being less’, unless strictly controlling the daily fieldwork record there is an enormous risk of generating contradictory data, such as in the continuous distinction of ditch re-cutting sequences; certainly such inconsistency occurs within Mucking’s archives. Indeed, producing this kind of record on a year-round basis (without seasonal gaps for checking), the in-field infrastructure required to successfully conduct such an ambitious programme would probably be beyond the resourcing of even most of today’s comparable-scale developer-funded projects, let alone with whatever meagre funding was then available.⁹

The quarry’s annual area-stripping ‘rhythm’ meant that, given the excavation intensity, they were often digging just in front of the machines. As is evident in Figure 1.5’s lower image, there is something absurd about excavating the ‘face-edge’ enclosures to such a degree when the remainder of the exposed field had not been tested and there could be no anticipation of what problems lie ahead (see also Fig. 1.6).¹⁰

As emphasised in a recent project review (Barford 2011; see also 1995), Mucking’s recording was perfectly reasonable at the time of the project’s inception. Yes, over the course of the 1970s, with the advent of single-context recording it began to appear anachronistic – especially the use of imperial measurement – but, nonetheless, it was a system. Once it is recognised that the site’s horizontal strata were ploughed-out, leaving only features cut into the gravel natural, then the use of specific grid coordinates to locate excavation interventions was a ‘workable’ basis of site enumeration. Admittedly, its application was not entirely consistent, so that postholes set within the interior of a roundhouse gully could, for example, be assigned their own number sequence. Yet, in reference to the plans, all features can eventually be distinguished.

Aside from the many thousands of site graphics, the basic record sits in the archive’s 363 notebooks (Figs. 1.11–12), which were generally – but, again, not always – local to specific site-grid areas. Fills were essentially dug by arbitrary spits, ideally 3–6", with finds separated by their level-depth (theoretically those from spits were meant to be subdivided according to major soilmark distinctions) and ‘exceptional’ finds individually plotted. The site archives include a nine-page long ‘Notes for Excavators’. As outlined therein, digging/recording procedures were meant to be standardised (though notebook recording was essentially undertaken by the area-supervisor alone) and accompanying cartoon pages depicted the range of expected features types and how, also, ditch sections were meant to be laid out (Fig. 1.10). While not particularly sophisticated, this also included a nomenclature of the site’s common soil-types. In practice, however, much of the deposit recording is fairly perfunctory (e.g. ‘black soil’ or ‘sandy fill’).

In conjunction with the quadrant/segment coordinate-system allocations, numbers were meant to be serially assigned to a range of feature/structure types, with graves, Grubenhäuser, kilns and wells specified. This is significant as it reflects an inherently hierarchical ‘type-based’ approach to the site’s archaeology generally. Indeed, if roundhouses, various post-built structures and barrows are added to this list, then in many respects these constituted the basic ‘building blocks’ by which its archaeology was addressed (see Fig. 1.20). Major compounds were specifically entitled (e.g. ‘Belgic Banjo Area’ or ‘North Enclosure’; Fig. 1.3), but generally ‘ditch dynamics’ received less attention than is now common; instead, it was discrete named feature entities/types that were highlighted.

As well as containing much sketch notation, a number of the notebooks include drawings of significant finds. Similarly, the original finds catalogues (by category-type) were also extensively sketch-illustrated. While the generic finds categories from each ‘intervention’ were often not enumerated, ‘featured’ finds were serially numbered and usually drawn (e.g. feature sherds such as rims and bases, metalwork, as well as special flints and fired clay pieces; see e.g. Fig. 4.72). These were compiled during the course of excavation and would have assured a high level of finds ‘feedback’. Indeed, as is evident in the site’s many interim statements, there clearly was quite a firm grasp of its basic sequence at any one time.

What, however, was missing from this is any kind of centralised indexing. Accordingly, while they would have known how many Grubenhäuser or wells had been excavated to date, there was no such specific accounting of the site’s postholes and pits (see Fig. 1.13 concerning the MPX’s ‘card system’). The same is also true of its finds, as there was never any control of, for example, the overall amount of pottery recovered (this in part dictated by the piecemeal dispersal of material to available specialists). Nor in the course of Mucking’s post-excavation programmes have such unified indexes ever been assembled; though there are a multitude of ‘type’ indexes and catalogues, in many cases – including the total number of features (c. 44,000 estimated) – gross-values remain ‘ballpark’ estimates to this day.¹¹

Appreciating this is crucial for understanding how the site’s post-excavation was approached. While surely also determined by funding contingencies (and ‘more’ would have always been intended), in the end it was a matter of a selective totality. Though all features were dug and recorded, the Joneses were evidently not overly concerned with postholes unassigned to structures or unphased pits (see, though, Fig. 1.19). Equally, a period’s ‘bulk’ pottery was not, at least initially, something to be prioritised, but only those key wares and types: things that ‘spoke’ and could be readily dated (though all finds were nevertheless collected and stored). There is nothing particularly surprising in this and large-scale assemblages invariably require ‘type markers’ to enable their sub-division and eventual specialist study. Such procedures would have been all the more necessary prior to the establishment of regional artefact chronologies and sequences, and were common in pre-1980s archaeology. Indeed, in reports of that era the total quantity of features excavated or sherds recovered are figures that are almost never supplied; instead, there is, for example, the illustrated feature-sherd catalogues. It was only in the later 1970s and early 1980s that the aims of such excavation practices – especially the loss of so much ploughsoil data that they entailed – was critiqued:

Despite the fact that most excavations have been obviously only partial, totality of examination has often been hailed as the ideal, to be abandoned only because of its impracticality … there has been a fashion recently for large-scale excavations of settlement sites, for example Mucking (Jones 1978[b] and 1974) or Gussage All Saints (Wainwright & Spratling 1973), in which totality has been claimed as the aim. Even if one accepts that such totality is both a logical and a practical possibility, both of which I seriously doubt, a consideration of such total excavations is most revealing. The strategy usually involves as a first stage the mechanical stripping of the topsoil, thus ignoring and destroying all the information contained therein, such as the spatial distribution of artefacts and their relationship to subsoil features. Already part of the total information has been lost. Furthermore, never has the entire contents of a British site been sieved to recover such items as small animal bones and seeds.¹² Total excavation, in fact, all too often involves nothing more than the total recovery of such traditionally determined features as houses and such artefacts as pottery (Champion 1978, 208; emphasis added; see also e.g. Barker 1980).

It can be questioned whether the Joneses were fully aware of just what they were taking on with the site’s intensive excavation and, especially, why they continuously dug so many kilometres of enclosure ditches. Within the project’s many interim reports, while certainly aware of the value of site distributions, it is clear that only once the computerisation of its archive was progressing at the Institute of Archaeology, London (see below) did Margaret began to fully realise the distributional possibilities of the project’s vast data-base:

Figure 1.11. The Mucking ‘Stacks’: left, the site notebooks (top) and, below, cumulative sources (photograph, D. Webb); right, notebook plotting. The individual area-coverage of all of the site’s notebooks has been plotted and collectively produces an image akin to a psychedelic plaid. For the purposes of comprehension, here only the coverage of seven notebooks (Nos 1, 50, 100, 150, 200, 250, 300 & 350) is shown. Although limited, the sample plot is both representative of the site’s progress and its various modes of recording. Whereas many are locally specific, with for example Notebooks 1 and 300 respectively dedicated to Roundhouses 1 and 97, Notebook 50’s coverage extended along a broad front and reflects the orientation of the quarry’s working face at the time. In contrast, Notebooks 100 and 150 were largely cemetery-related; the latter’s cover being markedly widespread.

The question – What use are such graphics in archaeology? – is best answered by another – How else could one quantify and describe the occurrence of very many artefacts except by mapping?

Distributions maps have been a mainstay of archaeological thinking especially since Cyril Fox published with Lily Chitty no less than forty one distributions in their classic Personality of Britain fifty years ago. But distributions within sites have rarely been tackled on a meaningful scale except by geophysicists’ recording surveys of anomalies. Since scientists handle such work computers are accepted as a matter of course.

Most archaeologists being still arts based, computers are still suspect …

If anything is going to re-create the past environment on this Thames side terrace gravel and enable the results to be published economically surely it must be this new data technology (Jones 1985b).

Figure 1.12. Site Records: top Tim Potter and Francis Pryor’s site notebook entries, Christmas 1965; below, BM/EH-phase context sheet.

Figure 1.13. Archival Curios: left, a concise explanation of the MPX record card system, with an example below; right, 1970 letter from Gen. Sir Charles Harrington, MoD’s Chief Adviser on Personnel and Logistics, to Rt. Hon. Sir Eric Fletcher, MP for Islington, concerning Margaret’s application for the army to machine-strip the site that year. With the card system note typed on the back of a used A4 envelope, together these ‘relics’ reflect two characteristics typical of Margaret: extreme thrift and audacity!

Problems arose with Mucking’s system essentially because it was based on a pre-Frere Report approach to site analysis/publication, whose impact is clear in a statement of Margaret Jones from 1980:

To add to the sheer physical problems of handling so much material from the 13 year excavations are the consequences of current revolutions in printing technology and mechanical data handling, as well as new thinking in archaeology which requires all excavation data to be made available. In the past, an excavation report consisted basically of the excavator’s narrative of what he had dug and his conclusions on its meaning. His argument was supported by selected plans, finds and specialist reports. Today’s excavation results need to be presented quite differently. The intention is to enable data to be re-assessed by others – now and in the future – who might well reach different conclusions. The principles of this new publication policy were debated by a quango whose recommendations are known from its chairman – Professor Frere of Oxford as the Frere Report (emphasis added; note her masculinisation of ‘the excavator’; Jones 1980).

There can be little doubt that the intensity with which the site was dug and the vast quantities of data it accrued (e.g. its estimated 1.7 million finds), was ultimately to the detriment of its post-excavation. This is not only true in terms of the ability to handle such a staggering amount of information, but also ‘the gap’ it inevitably created between what expectation there would have been of its final publication format and the interim awareness of its fieldwork results. With, for example, an entire volume intended to be devoted to the study of its fired clay, whether any coherent synthetic overview would have ever resulted from such overly in-depth study is debatable.¹³

At this juncture the history of the project’s post-excavation requires some rehearsal. Aside from Wainwright’s ‘Time Please’ retrospect of the era generally (2000; cf. Fowler 2001 and Rahtz 2001b; see also Cooper 2013 and Cooper & Yarrow 2012), this volume benefits from the recent appearance of Barford’s appropriately entitled appraisal, ‘Mucking: Real heritage heroism or heroic failure’, in The Great Excavations volume (Schofield 2010). In addition to sketching the site’s background and singular social milieu, the latter also outlines the tale of the previous two phases of its post-excavation aftermath. With the details therefore provided by him, and augmented by The Site Atlas’ summary (Clark 1993, 12–16), here we can limit ourselves to basics.

Altogether, significant post-excavation resources went into the project (equivalent to c. £2 million in today’s money). Over time, a lot of personalities were involved and many got bruised in the process; it is not our intention here to judge their efforts, but rather to try to understand what was attempted and why. When compared to the some £85,000 that went into the site’s excavation (excluding much direct support from Essex County Council), the c. £250,000 that was directed to its initial 1978–85 ‘Mucking Post-Excavation’ phase – hereafter MPX – was considerable (see, by way of comparison, Fig. 1.7’s caption for Dragonby’s costings). It involved the employment of six Department of Environment-funded posts and 10 part-time Manpower Services Commission staff (hereafter, respectively, DoE and MSC), who were housed in the Central Library and Museum at Grays, Thurrock. Work then progressed on preparing the archive (e.g. inking plans) and on the site’s main specialist analyses; also, various phase/area plans were assembled and a vast number of finds drawings were completed (many featuring in these volumes).¹⁴ Certainly among its most important outputs was the computerisation of its archive, which was undertaken by Jonathan Moffett, under the supervision of Ian Graham, at the Institute of Archaeology, London. Although inputting of some portions of the site’s finds catalogues were never fully completed (e.g. Roman surface pottery and a substantial proportion of the prehistoric wares), by 1983 they were to run a series of distribution plots that Margaret was able to proudly announce in the pages of Rescue News and Panorama (see Fig. 1.17 and, also, Figs. 4.49 and 5.15, 5.20 & 5.21; the project’s computing is further detailed in the Roman Vol.).¹⁵

Coordinating a post-excavation programme of this scale obviously involved great effort and Margaret was, after all, 62 when it commenced. It would seem, however, that its main shortcoming – aside from a rather naive faith that ‘the computer would sort everything out’ (Moffett 1989 and Barford 2011, 225) – was its ‘simultaneous-ness’. Too many fronts were advanced at the same time (e.g. finds publication drawings) without ever having the phasing totally sorted or the finds processed and, generally, while many facets were initiated, few tasks were actually completed. Although the demise of MPX was publicly accredited to Margaret’s retirement, it is clear that handing over control of the project was not her choice and that the DoE Backlog Working Party, who oversaw and funded the project, had simply lost faith in her.¹⁶

The decision was made to continue the project’s post-excavation at the British Museum, jointly managed by the Museum and English Heritage (hereafter, BM/EH phase; Clark 1993, 14–17). Under the direction of Ann Clark and operating out of the Museum’s Blythe Road out-station from 1986–9 (though continuing in a much reduced version until 1992), this is estimated to have cost c. £325,000. The programme had a dual remit. On the one hand, the Museum was insistent that the archive be fully ordered for registration/accessioning purposes; on the other hand, English Heritage had a determined publication emphasis. Relating to the latter, while successfully compiling The Site Atlas (Clark 1993), the Anglo-Saxon volumes came to command the available resources (Hamerow 1993 and Hirst & Clark 2009). This choice was logical insofar as it was possible to isolate more readily that period’s features from Mucking’s ‘mass’ and it also focussed upon what was widely held to be its landscape’s most significant occupation horizon (the selection also being driven by the Museum’s need to accession the site’s ‘treasure’ grave-goods of that date; A. Clark pers. comm.).¹⁷ The resultant publications were produced in a traditional manner that focused on exhaustive finds presentation – effectively at a level that would be impossible to sustain for the rest of the site’s periods.

Figure 1.14. Mucking’s ‘Phasing Wall’ and archive store, British Museum, Blythe Road out-station (photographs, D. Webb): opposite page, lower left, detail Belgic Banjo, Roundhouse 65 (et al.); above, close-up of the ‘Phasing Wall’ with the South Rings blued circles central and with the Linford plan attached upper left (see Fig. 1.8).

As to the archive, the Museum was simply unable to get the Institute’s computing running again. Although tempered in the knowledge that a high degree of MPX-phase data-input error had apparently been identified, they had to embark upon the computerisation of the site’s records and finds once again. To facilitate this, it was decided to generate single-context records for its main cut features. All told, some 6,000 such sheets were thus issued (see Fig. 1.12), but – restricted to features of more than a set size (> c. 0.60m across) – this was just a fraction of the site’s actual total.

As will be further outlined in this Chapter’s final section, here we have taken a different approach, relying, where possible, on the notebook record. The motivation behind the BM/EH programme nonetheless needs to be put in perspective. Still governed by the emphasis in the Frere/Cunliffe Reports on the archive, the 1980s saw the wholesale promotion of single-context recording systems, with notebook recording held to be an anathema (see e.g. Lucas 2001a). In reality, however, Mucking’s new context system essentially amounted to notebook cross-referencing/indexing. That said, their hand-drawn Atlas plans have proven a singular boon to our efforts.¹⁸

With the BM/EH programme’s resources hard-pressed between the poles of publication and never-ending archiving, selected study was made of two of its prehistoric enclosures. Steve Trow focussed upon the Iron Age North Enclosure. While primarily undertaken in order to check whether the context system would stand up to further analysis, as well in relationship to the original records, this also involved appraisal of its ceramics. Later, Jon Etté pulled together the records of the Late Bronze Age South Rings and analysed its ceramics, with the intention that the monument would feature in a Mucking prehistory publication. Although, due to funding constraints, the latter was not realised, our researches have greatly benefited from both of their analyses and draft texts. Equally, during the programme’s 1989–92 scaled down ‘aftermath’, Chris Going had a wider remit to examine the site’s Roman phases and produced a draft volume text containing much provisional specialist work, which similarly informed our ‘Mucking Roman’ volume.

Amidst its reconfigurations, a number of project specialists did not survive the ‘regime changes’ of Mucking’s protracted post-excavation. In hindsight, what does, though, seem singularly remarkable in all this