Current Archaeology normally features dirt archaeology, but archaeologists today often excavate archives as well – that is to say, they are engaged in digging into the archives in order to publish definitive accounts of past excavations. Here Christopher Evans and Sam Lucy give us an idea of the challenges they faced in completing the last of the Mucking excavation reports, some 50 years after the start of those legendary excavations.
Everything started with St Joseph’s publication of the South Rings cropmark in Antiquity in 1964. Originally thought to be a Neolithic henge (it eventually proved to be a Late Bronze Age ringwork), the site at Mucking, in southern Essex, clearly warranted either protection or thorough excavation. With the unstoppable Hoveringham Gravel Quarry poised to swallow the monument, protection was not an option.
Excavation commenced the following year within the southern end of the expanding gravel operations, and then ran continuously for the next 12 years (plus a brief return in 1978 when the Central Unit came to dig the North Ring’s ringwork) – not stopping even in winter. Mucking developed from a single-site investigation into an exhaustive landscape excavation that encompassed some 18ha at a time when a trench measuring 100 square metres was considered large. All told, this involved some 5,000 participants under the direction of the indomitable Margaret Jones and her husband Tom. It was rescue archaeology par excellence, and became the stuff of fieldwork legend.
Digging a landscape
Mucking was an excavation that grew organically, and its eventual scale was never envisaged at the outset. Starting a decade or so before single-context recording became standard practice, and long before the current era of digital capture, every layer, find, section, and plan was recorded by hand in one of 360 notebooks. It is not surprising, therefore, that Mucking’s protracted post-excavation analysis proved troublesome when this began in the 1980s – first, with a team employed at Thurrock Museum under the Manpower Services Scheme, and later with a dedicated team at the British Museum.
What emerged eventually was a site atlas and a brief summary volume. The Anglo-Saxon settlement and that period’s renowned cemeteries were published next, but, despite several attempts, the prehistoric and Romano-British phases remained locked in the archives. Aside from the South Rings and other numerous enclosures, this encompassed a truly enormous amount of archaeology, including more than 110 roundhouses, some 170 Roman-phase burials, and 23 contemporary kilns.
For us, everything began in 2006, when J D Hill at the British Museum approached the Cambridge Archaeological Unit to see if some manner of publication could be pulled together for these ‘missing’ periods. BM staff had recently been told that they would have to vacate the Blythe Road building where the site’s vast archives were housed. Many of the finds had never been accessioned and the unpublished material was at risk of being moved to a less accessible store and effectively forgotten.
After an initial archival assessment phase (funded by the late lamented Aggregates Levy Sustainability Fund), our programme was largely run ‘seasonally’ over an eight-year period, based on the availability of relatively small annual grants from the British Museum and the Society of Antiquaries of London (supplemented by oneoff contributions from other institutions, with Historic England funding the publication itself). Who knows whether we would have embarked on this if we had been aware from the start what it would eventually entail. The site’s last two volumes are now, nevertheless, at long last completed, and it is only fitting that the prehistory and summary volumes have just been issued in time to mark the site’s 50th anniversary.
Nothing is comparable to Mucking in British archaeology, with an estimated 44,000 features in total, including more than 400 structures and 1,145 burials. The story of the site begins with a succession of Early Neolithic, Grooved Ware, and Beaker-attributed occupations. Eight earlier Bronze Age barrows were found, plus a Middle Bronze Age field system with an accompanying settlement. It was, though, with the establishment of its two ringworks during the Late Bronze Age that the site, whose economy was fuelled by metalworking and salt production, begins to look different from other parts of the landscape, not least because of the continuous high density of occupation that stretches from the beginning of the first millennium BC through to the early Anglo-Saxon period.
All this is explicable when you take into account Mucking’s location on a prominent river-bend in the Thames estuary. A major insight into the importance of this location came from a completely unexpected source, when we stumbled upon the 1864 map of the reaches of the Lower Thames by Samuel (‘Self-Help’) Smiles; it was clear that Mucking’s terrace enjoys a unique viewshed straight down the river and its estuary – there being nothing between Mucking and the Continent. Add to this its dry and elevated situation on the terrace, its proximity to Tilbury’s river-ford, its access to arable hinterland, and to marsh and estuary resources (particularly salt), and it is obvious that Mucking was a distinctive place in the landscape, and an ideal site for settlement. All of this helps to explain the site’s extraordinary sequence.
The importance of Mucking’s Thames-side context was certainly appreciated by Margaret and Tom Jones with regards to the site’s Beaker and Anglo-Saxon phases. To their credit, they did not focus single-mindedly on the site’s potential long-range contacts: they also stressed the need to establish an over-arching regional finds sequence, with emphasis given to ‘the local’. This, for example, was expressed in the attention paid to its Romano-British kiln products. In our story, though, it is the site’s Continental imports that command centre stage, as they more clearly tell of the site’s ‘specificity’.
To enter the Mucking archives in the British Museum’s Blythe Road outstation felt like encountering some kind of ancient tomb. The opposing corridor walls were hung from floor to ceiling with the Jones’s original phasing plans, and these striking illustrations serve as introductory figures within the new prehistory volume. Passing endless finds-packed shelves, one eventually gained access to the main stores, with their great stacks of files, notebooks, and plans, evoking the final scenes in Orson Welles’ Citizen Kane (1941) and the film’s ultimate question: is this what a life amounts to? Even more absurdly, midway through our programme the finds were moved to the former naval arsenal tunnels at Dean Hill, Hampshire, where they are now guarded by plaster casts from antiquity, including the Knidos lion and the Parthenon’s caryatids.
It was in part in answer to that Citizen Kane question that we decided to give the ‘Lives in Land’ subtitle to these new volumes, as a reference to just how much of their lives Margaret and Tom Jones dedicated to Mucking. Equally, it resonates with the scale of Mucking’s dead and the hundreds of sand-stain images of its inhumations, there being very poor bone-survival given the area’s acidic soils. Almost like some kind of bizarre ‘dance of death’, these striking X-ray images attest to the sheer number of lives lived out on Mucking’s terrace. This sense of pathos and the mausoleum-like nature of the archive were further hammered home in the knowledge that many of the participants in the reams of dig-team photographs in the archive, and the thousands of names listed on large, annotated graph-paper sheet registers, have now passed from youth to old age, or have passed away themselves.
Mucking was undeniably a great social cauldron: indeed, for a generation of young archaeologists, it proved something of a rite of passage. We have included a series of recollections from some of its leading players (a number of whom later went on to be the great and good of the discipline), which – although not amounting to any kind of comprehensive oral history – provides a sense of its time and place to the newly published monographs. As well as critiquing the site’s methods and quasi-regimental organisation, these participants refer to certain themes and stories that appear to have imprinted themselves on those who worked at Mucking: the windswept site, the cold and primitive living conditions, the sorry site dog, Reject – and the occasion when Tom Jones scavenged hundreds of pounds of bananas from the Tilbury Docks and everyone lived on them for days on end. Yet, in most of these recollections, there is a sense of warmth and respect for what Tom and Margaret achieved. There is also considerable humour, with perhaps the best line coming from Warwick Rodwell: ‘what to say about someone [Margaret] who has a box in their hut labelled “pieces of string too small for any use”’.
Given the paucity of available resources, the Joneses were certainly audacious in what they attempted. Propelled by the rescue-led ethos that total site destruction warranted a total excavation response, their goal was to excavate every single feature in its entirety. Though, in truth, this generally amounted to around 75%, ditches were still dug in their entirety over several kilometres. The sampling theory developed by the Young Turks of the later 1970s brought with it criticism of the Mucking approach on the grounds that ‘total excavation’ only involved conventional categories of finds, and that it lacked sufficient sieving of deposits or acknowledgement of topsoil finds. In fact, both of these were attempted, and the site’s intensive excavation strategy in effect acted as a ploughsoil sampling grid, because of the recovery of great quantities of residual finds, particularly flints. This, plus the cumulative density of its features, permitted us a much higher level of successive pre-Middle Bronze Age ‘settlement/ use cluster’ analyses than most of today’s massscale, low excavation-intensity sites.
What undermines the site’s espoused ‘totality’ more significantly was quarry-track disturbance and the machining of large areas of the site. Again, it is not a matter for finger-pointing: this was the infancy of machine-stripping projects, and the archaeologists did not always have complete control of the way the site was graded ahead of mineral extraction work. It is estimated that approximately one fifth of the site was disturbed in this way, especially towards the southern end, and saw little or no recording as a consequence. This obviously has an impact on the site’s finds distributions, and shallow-footed building remains were even probably lost there as a result. The Manpower Services Commission (MSC) phase of the project resulted in a series of computer files that could not be unlocked in the British Museum phase of post-excavation work. Fortunately, with the more sophisticated datasalvage techniques that are now available, we were able to get these files operational again (albeit at a cost of considerable time and effort). Suddenly we had more than a million grid-referenced finds at our disposal, and their distributional plotting became the main backbone of our programme.
Margaret herself had a background in geography, and was clearly well aware of the value of spatial distributions. Within the archives there is a series of plastic sheet-printed distribution plots, which evidently derive from the site’s initial computer records, developed in collaboration with computing experts at the UCL Institute of Archaeology. The records show that the site’s publication was originally going to include a standard-size portfolio of these, which were meant to be draped over a base-plan.
It can be argued that the Joneses were, in fact, over-ambitious in what they attempted. When the fieldwork commenced, not only was computing not easily available, neither, to all intents and purposes, was photocopy-reduction. Plans had to be scaled down by various mechanical means and, thereafter, reproduced by photography or as dye-lines. As a result, their scope for revision and updating was limited, and the archives abound with taped extensions to plans (and numerous versions of texts). Indeed, it includes what must be one of the most extraordinary plans in British archaeology – one that certainly deserves inclusion (see p.32). Of MSCvintage, it shows the many thousands of unphased features at Mucking – largely undated postholes. It must have required many thousands of personnel hours to assemble, and as such it is an amazingly frank admission that there was much about the site that could not be dated or brought within an overall interpretative scheme.
This figure is but one of a series of period phaseplans that were then drawn up, and we estimate that they got somewhere in the region of 80% of the site’s sequence ‘right’. This raises the issue of whether what we have been able to do was worth the 30-year wait, and whether it has added anything substantial: it thereby reflects on debates about the role of ‘understanding’ in archaeology versus data, evidence, and ‘proof’.
When Mucking began in the 1960s, the goal of publication was essentially to present selected material illustrative of a site’s sequence. With the issuing of the Frere Report in 1975 (and, eight years later, the Cunliffe Report), the emphasis of postexcavation work shifted to the preparation and availability of the full archive (usually achieved by publishing the supporting data on microfiche as an accompaniment to the printed interpretation). This latter approach clearly posed major pragmatic problems for Mucking, and these were never entirely resolved – for example, the archive still lacks an index of master features. In short, Mucking’s post-excavation analysis and publication fell foul of changing trends in archaeological reportage and, with justification, it could be argued that the site should have simply been issued as a 1960s-era publication. Yet, without this change of post-excavation philosophy, the finds would never have been computerised and we would not have been able to build on this archival legacy by analysing the distributions of finds and features.
A major shortcoming underlying our work is that not all of Mucking’s prehistoric pottery was studied, and for the Iron Age phases we have had to rely on a sample from its main enclosures and roundhouses. That period’s representation has also been plagued by other problems. It took time to realise that most of the site’s Late Iron Age wares had actually become incorporated into the Roman ceramics database, and it was only by enlisting the good graces of Isobel Thompson that this could be rectified.
A more serious issue related to one of the project’s ‘road to Damascus’ episodes. Based on the state of knowledge of the local ceramics, the South Rings site was originally thought to have been an Early Iron Age ‘mini-hillfort’. Yet, with John Barrett’s analysis of Post-Deverel Rimbury Wares and their reassignment to the Late Bronze Age – coming as something of a seismic shock to mid-1970s archaeology and featuring heavily in that decade’s BAR published conference proceedings (as well as in the pages of this magazine) – the Rings were suddenly backdated by some centuries.
This reappraisal had the major knock-on effect that, with almost all of the features attributed to the Early Iron Age now reassigned to the preceding period, we were left with a marked hiatus in the sequence. Given our limited resources, it was not feasible to re-examine all of the site’s Late Bronze Age/Early Iron Age wares (especially as they were then stored away in the Dean Hill tunnels). A pragmatic solution presented itself in that, during the creation of the MSC catalogue, literally thousands of these prehistoric sherds had been drawn. Through reference to these, Matt Brudenell was able to achieve a solid sense of the distributions of the respective wares, and to identify a number of major pottery groups from both periods.
Having to divide Mucking’s sequence to fit into separate books involves the kind of ‘timeslicing’ that invariably generates problems, not least because it is at the interfaces of the main periods that the programme’s major discoveries were made. Perhaps the most important of these was to confirm an original Jones-era interpretation: Mucking really did see transitional Roman to Anglo-Saxon occupation in the first decades of the 5th century AD, perhaps even in the latter years of the 4th – as is evident from the fact that the site’s later Roman pottery largely only occurs within the earliest Anglo-Saxon Grubenhäuser at the site’s southern end.
Our programme’s other keynote contribution was the recognition of a major Late Iron Age ceremonial complex within the site’s centre, which rapidly became known by our team as ‘The Plaza’. Along with the South Rings, this is one of the pivots of the site’s sequence. Lines of cremation-related square barrows defined the short sides of this quasi-rectangular setting. The conjoining western axis consisted of a streetlike array of four- and six-poster raised granaries, plus a nine-poster at one end of what we named ‘Granary Row’. In the Conquest Period, the sides of The Plaza were delineated by fencelines, and thereafter it remained essentially an ‘empty’ or reserved space.
The Joneses saw these fences as relating to subsequent horticultural activity and (at least originally) they thought that the barrows were ‘Belgic houses’ that had cremations set within their interiors at a later date. Now, though, with our greater familiarity with Late Iron Age layouts, we can say without doubt that The Plaza functioned as a ceremonial ‘whole’ and as an expression of social power. Its linkages may have been with the near-Continent, because the only known direct parallel for the linear arrangements of Mucking’s square barrows is in the Champagne district of France – a point reinforced by the occurrence of Gaulish Terra Rubra pottery on the site. The underlying basis of this connection may have been grain export. It is estimated that the storage capacity of the line of raised granary settings would have been sufficient to feed upwards of 150 families, far in excess of the ten or so households thought to have then been resident on the terrace.
It was the existence of Mucking’s Late Iron Age activity that evidently determined the succeeding Romano-British estate centre and its appended settlement. An overseer’s residence has been identified with an incredible highstatus assemblage backfilled into an adjacent 15m-deep well, as well as rich burials and a major granary. Aside from pottery production – some of its products being destined for the military on Hadrian’s Wall – it was essentially an agricultural powerhouse. In fact, grain and salt were the long-term ‘drivers’ of Mucking’s economy from the Middle Bronze Age period onwards.
One of the challenges of writing the final Mucking volumes has been striking an appropriate balance between these being our books, but also the Joneses’. We continually asked ourselves ‘how would the Joneses interpret this or that’, and we were helped in this by their numerous interim reports. Even so, their concerns cannot restrict ours. We cannot really envisage what it would have been like to do archaeology at a time when there was so little excavation precedent – for instance, very few Iron Age or Romano-British settlements had been exposed at any scale when Mucking was under way. As we have come to know more about both periods, it is clear that some of the site’s components are no longer particularly ‘special’ in themselves. On the other hand, there can be no ‘Mucking for the 21st century’, because the excavations remain intrinsically rooted in their time, and though our emphases may differ, the archive cannot somehow be updated or made better.
For these and other reasons, it is appropriate that the last two Mucking volumes appear within the Cambridge Archaeological Unit’s ‘Historiography and Fieldwork’ series. In essence, we’ve drawn a line between them and the earlier Mucking publications. We no longer consider them as pertaining to a current-era site and, instead, they are a matter of ‘historiographical rescue’. Like our Fengate Revisited volume of 2009, they are intended to have a ‘raw’ quality and be more scrapbook-like, and not overly concerned with the niceties of uniformity of graphical style, for example (but then, given the restriction of our funding, this was never a realistic option when faced with the wealth of the site’s archive drawings).
The two volumes have involved more than 50 contributing specialists. A number have, in effect, been reactivated from the original team and, with many participating gratis, there has been a real sense of a community working together to finally get the site ‘put to bed’. While there have been times of immense frustration, it has only been a great privilege to be able to work on the huge data-sets that the Joneses generated. In the end, we think we have granted the Joneses a measure of justice for their sheer perseverance and dogged determination on that Thames-side terrace. Certainly it is difficult to think that there will ever again be an excavation like that again: what they achieved there was something truly heroic.
IMAGES: T Jones