The mystery in the marsh: Exploring an Anglo-Saxon island at Little Carlton

12 mins read

In May 2014, Current Archaeology reported on the discovery of a plaque inscribed with the name of an Anglo-Saxon woman, ‘Cudburg’, at Little Carlton near Louth, Lincolnshire. The site has since emerged as one of the most important high-status settlements yet found in the region. Peter Townend, Hugh Willmott, Adam Daubney, and Graham Vickers explain how they brought to light the story of an early medieval marsh-island community.

Metal-detector and geophysical surveys, LiDAR scans, and archaeological excavation have all come together to reveal a previously unknown high-status site at Little Carlton, near Louth.

If you visit Little Carlton – a typically tranquil rural settlement in the Lindsey Marsh, where traces of medieval ridge and furrow, paddocks, isolated farms, and historic buildings nestle between pockets of woodland, arable fields, and grassland – it is easy to feel that little has changed for centuries. Certainly, the Lincolnshire parish, which lies 9km from the east coast, is known to date back at least as far as the time of the 1086 Domesday Book, in which it is listed – while the discovery of a Late Saxon (later 10th or 11th century) grave cover during the demolition of the local church provided further hints of early activity. Little else was understood about this chapter of the parish’s past, however, until metal-detectorist Graham Vickers brought an intriguing recent discovery to Adam Daubney, Lincolnshire’s Finds Liaison Officer (FLO), for recording under the Portable Antiquities Scheme (PAS) in October 2011.

Finds Liaison Officer Adam Daubney and metal-detectorist Graham Vickers worked together to record dozens of Middle Saxon metal finds from the site before excavations even began.

The object was a silver stylus, an ornate writing tool. Although it had been recovered from disturbed ploughsoil, its distinctive decoration, incorporating ‘Mercian-style’ animal and chip-carved interlacing, placed it squarely in the 8th century. This stray find held tantalising clues to a potentially significant settlement lying somewhere nearby: not only did it speak of Anglo-Saxon inhabitants who could read and write, but the fact that similar items have only previously been found at high-status sites, such as Flixborough, also in Lincolnshire, was another promising indication. This was no ordinary find – and it would only be the first of many. Over the next few years, Graham and Adam recorded many more metal finds from the plough-zone, most of which, like the stylus, could be dated to the Middle Saxon period (c.AD 710-850). Each was a new piece in an increasingly exciting jigsaw puzzle, complemented by the several thousand sherds of both Ipswich ware and Continental ceramics, and items both domestic and luxurious – from whetstones and loom-weights to fragments of glass – that were also found. The richness of this assemblage was matched by Graham’s professionalism: each discovery was plotted using GPS and promptly reported to the PAS, and in a short time he had amassed a data set that was unique in both its size and recording precision. It quickly became clear that an important new site had been discovered. But what would this huge mass of material reveal?

Styli and stylish finds

Many of the Little Carlton finds were characteristic of a typical Middle Saxon settlement: domestic and personal items, from strap-ends and pins to hooked-tags and tweezers. But above all, the assemblage befits a community who enjoyed and had access to the finer things in life. Some of the artefacts were distinctly luxurious, notably a stunning glass counter decorated with a number of colourful twisted strands, which may have been set in a bronze bowl.

Equally unusual was the number of styli found here – 16 in total – which provide clear evidence for literate people engaged in copying or composing secular or religious texts. Meanwhile, another piece of evidence for literacy had a more personal flavour: a small lead tablet bearing faint but legible letters spelling ‘Cudburg’ – a female name – interspersed with small crosses (CA 291). Analysis of the lettering by Professor Elizabeth Okasha of University College Cork revealed that it was incised using ‘insular majuscule’ script (written entirely in capital letters) with neat serifs, suggesting a pre-10th-century date and an author used to writing on vellum. But what was it for? Similar fragments from Norfolk and north Yorkshire have been interpreted as funerary objects, possibly name plaques to be placed inside a coffin. Might Cudburg have been one of the inhabitants of our Middle Saxon settlement?

An unusual find from the project: a glass counter decorated with twisted colourful strands.

The lifespan of the site itself could also be read in Graham’s finds: among them were a number of coins, mostly sceattas spanning c.AD 680-790, though there were also several broad pennies testifying to the site having been occupied into the third quarter of the 9th century. There, however, the sequence ended abruptly – and it is perhaps no coincidence that the late 9th century was also a time of major Viking incursions into the region. Might these events have spelled the end for what seems to have been a temptingly wealthy and apparently undefended site?

In spite of the wealth of information provided by Graham’s finds, major questions remained. Metal-detector assemblages are useful as a general indicator of a site’s likely chronology and status, but large bodies of material, like that from Little Carlton, are notoriously difficult to interpret. For example, metaldetecting finds could simply indicate periods when metalwork was most common at a site, which might not be representative of the broader picture of its material culture. In addition, most objects tend to be fairly vague in terms of the dates they indicate, and for the most part we cannot tell whether batches of artefacts were deposited in one single ‘episode’ or over longer periods of time.

How far can we truly identify a ‘high-status’ site from its finds? Might there be far more lowerstatus individuals living there than higher-status ones, who do not show up in the material record? Finally – and perhaps crucially – it is often difficult to establish a site’s purpose through its artefacts alone. Archaeologists have traditionally been quick to assign sites to distinct categories – monasteries, fairs, markets, and so on – but the reality is that sites were not static, and their status, ownership, and function could all change within the lifetime of a single object. What we needed to do next was put the finds in context by examining the landscape in which they were found.

What’s in a name?

Our first step towards exploring the wider landscape was to create a detailed plan to plot the distribution of Graham’s finds. Using this, we could see a clear cluster of Middle Saxon material tucked tightly where the medieval parish church of St Edith once stood, while the fork of two rivers marked the limit of finds to the north. Artefact numbers diminished equally rapidly to the south, once we moved beyond the road. As we explored these patterns further, several unusual (and, indeed, previously overlooked) aspects of the landscape began to make more sense.

Of all these patterns, the most distinctive was the way that Graham’s finds dropped off as he moved further from the road. Our survey of the landscape revealed that it was not only the number of artefacts that was becoming lower as we looked south, but also the level of the land itself. Might the two be linked? Analysis of the names of these ‘unproductive’ southern fields, as recorded on 19th-century maps, suggests that they might be. In 1820, these areas were known as ‘Little Fen’ and ‘Horse Fen’, suggesting lowlying, marshy land unsuited to settlement. This might explain why we found no sign of occupation here – even though today the fields are dry, having been drained by post-medieval farmers to reclaim the land for agricultural use. Sure enough, in the area to the north, at the meeting of two rivers where finds also petered out, we found another marshy field-name: ‘Engine Fen’. By contrast, the land that these fields surround – the land that saw the focus of our finds, and later housed the parish church and post-medieval manor house – was noticeably higher, which must have once created a habitable island rising out of the medieval marsh.

At the heart of this raised ground, where the bulk of Graham’s discoveries were made, another field-name leaped out: ‘The Bruff’. This probably reflects a local pronunciation of the Old English word burh, a ‘defended site’. We see it in various guises scattered along the Trent – Burton upon Stather, Flixborough, Gainsborough, Gate Burton, and so on – where it is thought to possibly indicate the presence of Middle Saxon forts controlling landing places along the river. Might our site have been similarly strategically sited? Certainly Little Carlton is conveniently nestled between two waterways, the Beck and the Old Eau, which merge to form the Long Eau, a meandering river that empties into the North Sea at Saltfleet Haven some 12km to the northeast. That Saltfleet was an important early medieval port is attested by the Domesday Book, as the disputes section of the 11th-century survey preserves a complaint about ships there being taxed ‘like never before’. Could Little Carlton have been deliberately located, and have flourished, because of its ability to control coastal trade? This is an exciting line of enquiry that future investigations might yet answer.

Moreover, Little Carlton’s Middle Saxon inhabitants do not seem to have been the first to recognise the site’s advantages. Among Graham’s finds there was also a small and rather eclectic group of Roman items, spanning all four centuries of Roman occupation in Britain, though in very small quantities. These ephemeral finds represent a halo scatter of activity, suggesting more the periphery of a Roman site than a settlement itself. Might we be on the verge of locating even earlier occupation at our site? Or might these objects be treasured heirlooms curated into the Early Medieval period? The presence of a 6thcentury Anglo-Saxon brooch provided further tantalising hints of continuity – but this was a matter that only excavation could resolve.

Surveying the site

Moving our project forward was not as simple a matter as just saying that fieldwork was required, however. Although the PAS has made great strides in recording finds, logging its millionth in September 2014 (CA 297), few such assemblages have been investigated archaeologically. Excavation is costly, and unfortunately the PAS does not currently have the resources to fund research. Any attempt at such an investigation, then, relies on a proactive approach from the local FLO, the landowner, and the finder, as well as anyone else who might be in a position to help. We were lucky in this respect.

Last summer, Peter Townend and Hugh Willmott from the University of Sheffield’s Department of Archaeology agreed to take time out from their project at Thornton Abbey in North Lincolnshire to carry out a targeted geophysical survey of our site, focusing on the fields immediately to the north and south of where the church had stood, to see if they could locate any trace of a settlement associated with our cluster of finds.

There were no guarantees that such a search would succeed: given that the site had been extensively ploughed over the last few hundred years, it was quite possible that no identifiable archaeology would survive below the surface. Despite this, Peter and Hugh’s magnetometry survey was unexpectedly clear, revealing the distinctive lines of boundary and drainage ditches, and possibly even the foundations of buildings. There were further hints of earlier activity too, with a circular ring-ditch identified to the east interpreted as a probable prehistoric barrow – though this was not investigated due to the limited scope of the excavation.

Analysing the distribution of metal finds showed that they formed a distinctive cluster north of the parish church, and petered out south of the modern road.
Analysing the distribution of metal finds showed that they formed a distinctive cluster north of the parish church, and petered out south of the modern road.


Interestingly, while the area lying outside the boundary ditch to the south was impressively busy, outside its northern limits there seemed to be no features at all. Why? To try to answer this, we produced a 3D model of the land surface using LiDAR data, which allowed us to visualise the landscape on a large scale, and to study it in detail. In this imaging, our island was much more obvious than it is today, rising out of its lower surrounds. To complete the picture, we raised the water level digitally to bring it back up to its early medieval height, based on the topography, geophysical survey, and distribution of Anglo-Saxon finds. This made all the difference to understanding our settlement’s original setting: it was now clear that the outer boundary ditch that we had located ran just within the limit of the surrounding water, but on the southern side of the modern road there was a distinct space of 10-15m between ditch and water, which was filled with a number of features. These results were illuminating, but it was increasingly evident that little more information could be gained without targeting a number of these finds for excavation. It was time to put spade to ground.

Digging begins

The next phase of our investigation saw us teaming up with student volunteers from the University of Sheffield to open nine evaluation trenches exploring selected features. Seven of these were located in the busy area north of the church: two over the boundary ditch itself – one to the east and one to the west; one over the dogleg of a central north–south ditch; two larger trenches at the intersection between a number of linear features, hoping to characterise as many as possible; and two inside the boundary ditch to explore some of the features we had identified within its confines. On top of this, we put in another trench outside of the ditch to compare this area with its interior, while in the field to the south of the church, a single trench investigated the relationships between the boundary ditch, the raised water-level, and the features we had identified in the space between the two.

These trenches revealed a wealth of information about the life of our settlement: many of the ditches had been ultimately used as refuse dumps, yielding quantities of Middle Saxon pottery and butchered animal bone, while our trench in the field to the south of the church more than delivered on its aim to reveal the settlement’s relationship with the water’s edge – an occasionally fraught one, it seems, as here we uncovered the remains of the community’s flood defences. A sturdy bank had been built up to guard against encroaching waters, while in an adjacent ditch we found the ends of waterlogged wooden stakes, driven into the ditch’s base and outside edge to revet the bank, probably once retained by wooden planking attached to the timbers.

Signs of settlement: excavation revealed tangible traces of early medieval occupation, including quantities of butchered animal bone.

Over in the trench dug outside the southern boundary ditch, the area between ditch and marshland was giving up extensive evidence of industrial activity, mostly metalworking. Most striking was a circular patch of burnt clay some 1.2m wide – the base of a hearth. Unfortunately, this feature had been heavily disturbed by later ploughing, which had destroyed its top, but we were still able to recover a large number of molten drops of lead and lead slag: signs of smelting. Given the noise and smell that such processes would have created, it is not surprising that they were limited to the marshy outskirts of the settlement. There were more finds to come at the southernmost end of this trench, where the land fell away sharply at the exact point where our 3D model suggests that the Anglo-Saxon era waterline would have been. Here, running parallel to the trench, were a line of postholes that may be vestiges of a quay or jetty from which the island community could have navigated the marsh. Piece by piece, we were coming to understand how they may have lived.

In the field to the north of the church, our investigations were proving particularly fruitful, not only confirming the findings of the geophysical survey in this area, but also yielding new information, revealing a not inconsiderable number of gullies, ditches, and postholes that had been too small for the magnetometer to pick up. All of these we could date solidly to the Anglo-Saxon period thanks to the fragments of pottery that they contained.

Two of these trenches were of particular note: the larger trenches that we had opened over a series of intersecting linear features. Cutting into the natural soil at their base were a number of gullies, postholes, and associated metalled surfaces: the well-preserved footprints of early medieval buildings. Although the restricted size of our trenches prevented us from reconstructing their full extent, the butchered animal bone and pot sherds found in the dark, rich layer sealing the structures allowed us to date them securely. Meanwhile, Graham – who had been working closely with the Sheffield students throughout the excavation – had discovered a copper-alloy  dress pin and a perfectly preserved sceat dated AD 725-745. Our wider landscape analysis had shed light on how the Middle Saxon community was able to thrive, and why such a concentration of artefacts from this period had been found around this spot, but now – at last – we had a tangible link to the inhabitants of this community.

Settling the site

Have we, then, solved the mystery of the marsh? Thanks to the impressive range of finds diligently collected and recorded by Graham, and our subsequent work to establish their context, we can now certainly say we know the total extent of the site, and that it was occupied probably continuously between c.AD 680 and 850, coming to an end at around the same time as the Viking invasions commenced. Despite its apparently sudden demise, Little Carlton seems to have been a ‘persistent’ place during its lifespan, carefully constructed around, and perhaps stimulated by, earlier forms of land use. If the circular ring-ditch proves to be of Bronze Age date, it may provide further evidence for the reuse of burial mounds in the Anglo-Saxon period, when such monuments often attracted ‘secondary burials’ (as at Barrow Clump on Salisbury Plain, see CA 306).

With the site’s clear evidence for occupation, industrial activity, and high-status activities such as writing and commerce, it is tempting to suggest that we might have discovered a previously unknown monastic or trading centre, but our work so far has only revealed a glimpse of this settlement – though it is evident that the site still has plenty of potential. Nevertheless, it is already possible to say that Little Carlton is one of the most important Middle Saxon sites to have been discovered in recent years, and one where the exciting research possibilities of cooperation between responsible metal-detectorists, the PAS, and university archaeologists are eloquently expressed.

This article appeared in CA 313.  Click here to subscribe!

Images: University of Sheffield; Portable Antiquities Scheme

Leave a Reply

Your email address will not be published.