Beyond the Walls: In Search of the Celts

Last month, Howard appeared on TV. Having been filmed on location at the Pillar of Eliseg and at Offa’s Dyke near Ruabon and looking across at Offa’s Dyke where it hits the River Dee at Ty Mawr Country Park in October, only a small section of the interview was aired, filmed on the dyke at Ruabon.

The documentary was produced by AlleyCats for BBC 4, entitled: Beyond the Walls: In Search of the Celts.

The programme explored the conflicts and interactions between the Romans and, subsequently, the Anglo-Saxon kingdoms, with the Celtic (Brythonic)-speaking peoples of the island of Britain. It was first air at 10pm on Tuesday 4th December on BBC4: The official blurb reads:

Historian Dr Eleanor Barraclough travels through some of Britain’s most beautiful landscapes – Hadrian’s Wall, the Lake District and Offa’s Dyke – in search of new evidence to reveal the true story of the mysterious ancient British tribes often called the Celts.

According to the official history books, the Celts were defeated and pushed to the edges of Britain by waves of Roman and Anglo Saxon invaders. However, a growing body of evidence suggests this is not the full story.

To help give the Celts back their proper place in our history, Eleanor examines freshly discovered treasures, new archaeological evidence from real photographs and clues hidden in ancient poetry to reveal a fresh narrative – one that suggests the relationship between our ancient British ancestors and those who came to conquer them was much less repressive, and far more co-operative, than we have thought.

I was being filmed in order to talk about the early medieval kingdoms of Britain, focusing on the relationship between the Britons and the Anglo-Saxons. To do this, I wanted to talk about the Pillar of Eliseg and Offa’s Dyke, and how they afford contrasting perspectives, one ‘Welsh’, one ‘Saxon’, on the Christian kingdoms of the Early Middle Ages.

me: whilst filming by Offa’s Dyke, near Ruabon

I won’t evaluate the programme, but make of it what you will.

All I can say is that I did my very best to articulate the stories of two complex monuments, and they gave me every opportunity to make my points as clear and coherent as possible. Whether I do make sense remains my responsibility, not theirs, despite being heavily edited! Certainly, this was my first time working on a TV documentary in a decade and it was a very positive experience for me to get to work with David Starkey and Alfie Turner from AlleyCats.

In doing so, I got to represent my work on Project Eliseg and my work with the Offa’s Dyke Collaboratory. even though I’m very disappointed that the Pillar of Eliseg didn’t make the final cut. Still, I was delighted to be able to talk about two monuments that should be regarded as key to the study of early medieval Britain, and yet which rarely feature on television documentaries.

The choice not to discuss the Pillar of Eliseg was odd. I had discussed some ideas to appear in the forthcoming Project Eliseg monograph. I addressed the complex biography of the early ninth-century cross, including its form and text. I also considered how the monument operated in its landscape setting, and how and why the monument reused a prehistoric cairn. Most importantly, I identified how the monument was positioned in relation to, and perhaps in defiance to, the Mercian linear earthworks to its east.


The Pillar of Eliseg in March of 2018

Regarding Offa’s Dyke, I addressed how it was positioned in relation to the Vale of Llangollen and the River Dee. I then talked about changing views on its functions and significance as a part of the Mercian frontier, and its legacy to this day.

What I hope the programme shows, even though the Pillar and the Dyke should/could have been discussed together, is the complex and fluid relationships between the British and the Anglo-Saxon kingdoms: a ‘prehistory’ to the March of Wales as it developed following the Norman conquest of England.

It was a pity that the equally significant Wat’s Dyke couldn’t be featured, especially in relation to the Pillar of Eliseg. Still, for a 29-minute programme, I think packed in quite a lot of new and interesting research! Many have commented to me that, even for abbreviated TV shows, it seemed far too rushed.

Cameraman/Director/Producer David Starkey, at work at Ruabon
David, busy trying to film the Dee Valley at Ty Mawr Country Park, and attempting to avoid the llamas…

Wat’s Dyke at Pandy off Plas Acton Road

Adapted from Howard’s Archaeodeath blog.

Around Wrexham, Wat’s Dyke remains an important monument but heavily damaged by agricultural, industrial and construction activities. Running c. 38 miles from Basingwerk to Maesbury, it is a monument often overshadowed by its larger neighbour: Offa’s Dyke. Excavations at Gobowen suggest that it is later than Offa’s Dyke: a Mercian construction of the early 9th century AD.

At Pandy, as part of a recent field trip exploring the industrial and post-industrial uses of the landscape, I encountered a section that runs from near the railway line across the fields towards what is now the A483, before it enters into what is now the suburbs of Wrexham. See my previous discussions of it at this location here, and nearby at Erddig here and here.

Fox, writing in Archaeologia Cambrensis in 1934, has this to say about the section.

Wat’s Dyke begins again 50 yards to the S. of the “Y” road junction, near Pandy Farm. (From this point northwards to the road, there is a mere hedge bank). The construction of the railway obscured the topography, but it is clear that the terminal point is exactly what was the head of a lateral (and doubtless boggy and thickly wooded) ravine opening northwards down Wilderness Wood to the main river (see Fig. 26, 5, a). From the very beginning, the Dyke was straightly aligned (Type I) to the 300ft contour, 620 years away, and was throughout on a large scale with a broad W. ditch present-day irregularities are due to destruction and levelling.

While the bank is denuded, it is clearly larger than any later field boundary, and straight-aligned with extensive views westwards. Moreover, nothing more than a wide depression running west of the bank, it is evident that a large ditch once existed on the dyke at this location.

In this position, Wat’s Dyke is not following the top of a ridge or ravine. Instead, it is running as straight as possible over relatively flat ground in order to navigate between the Alyn and the Clywedog rivers. To the south, it crosses the Clywedog before following the eastern ridge of the Black Brook past Erddig Hall. To the north, it crosses the Alyn and incorporates the earthworks of Caer Alyn hillfort.

We also encountered an inquisitive horse!


Harold’s Stones, Trellech

This isn’t precisely about Offa’s Dyke, but it is some of Howard’s musings about a site not far west of the Dyke, reposted and adapted from Archaeodeath.

I want to bring things down a key with this next post. I focus on the archaeology and folklore of an archaeological site I visited when in South-East Wales last year: Harold’s Stones, Trellech, Monmouthshire.

DSC03026These three megalithic stones stand in a line south-west of medieval and modern settlement of Trelech, just east of a stream and close to springs. This rare stone arrangement is intriguing and evocative: I’ve only previously addressed anprehistoric standing stone alignment once before for the much-larger and widely spaced Devil’s Arrows. The Trelech stones tip at different angles on a gradual hillside, the middle and north tip slightly westwards, while the tallest and southernmost has a striking lean eastwards. Indeed, Elizabeth Whittle (1992) described their collective impression as being ‘drunken’.

DSC03021The stones are set on a NE-SW alignment. Each stone is over 2m tall: with the largest to the south. They are each comprised of conglomerate puddingstone.

DSC03017The central stone might have been shaped by human hand, and has two cup marks on its south side.


Coflein tells me that geophysical survey around the stone revealed traces of medieval or later structures and a c. 40m diameter subrectangular ditched enclosure (presumably undated). Further images can be found on the Megalithic Portal.


So these are prehistoric standing stones, presumably of late Neolithic/Early Bronze Age date, and perhaps part of a longer alignment and other monuments, now lost. I recognise the standard later association with the Devil in folklore: namely that they were thrown from the Sugar Loaf mountain by Jack o-Kent – a giant, when playing pitch and toss with the Devil. Such satanic associations tend to be of early modern in origins.

DSC03006Naming: Why Harold?

Yet there seems to have been more to these stones in the medieval mind than devilish connections. The place-name ‘Trelech’ means the ‘settlement/village of the stones’. This suggests that the vicinity was defined by these prominent triad of megaliths that acquired the roles of prominent landmarks and perhaps also the association with rich now-lost fables.

Why Harold’s stones? Before getting one in the eye, Earl Harold Godwinson of Wessex and then Hereford too, defeated Gruffyd ap Llywyelyn in 1063 and thus continued his rise to power as a war leader. At some unknown point, these three stones could have been ascribed to Harold’s victories in battle, and more specifically, that each stone commemorates a chieftain slain/defeated in battle by Harold. I’m still intrigued by the possible and specific association of the Trelech stones with the man to briefly become King of England and died at Hastings in 1066.

Now, there are 3 “Harold Stone” place-names in Pembrokeshire, each associated with a single megalith:

All are single isolated standing stones, two in coastal locations, the third not far from the sea. Given the far-west location of this cluster, their association with the victories of Harold over the Welsh in the 1060s doesn’t seem very likely. However, they might well be connected to the travels of Harold and his fame. The logic of this cluster might lie less in 11th-century events and more than the extensive settlement of Normans, Saxons and Flemings during the 12th/13th centuries. Could these groups have been prompted to attribute ancient stones in the west Welsh landscape with earlier invaders of famed reputation?

Might such an explanation work for Trelech? Certainly, but perhaps there’s more. The triad of stones at Trellech is more logically located in relation to sites of potential Saxon-Welsh conflict. Trelech is located just west of the Wye and thus the line of the eighth-century linear earthwork – Offa’s Dyke. Hence, the feasibility that this is the site of a battle, or became equated with a series of epic conflicts between the West Saxons and the Welsh, is persuasive. Moreover, using the Pillar of Eliseg as an analogy, perhaps the stones served as a regular assembly place before and after such conflicts. What is perhaps important from a Welsh perspective is that any memorials attributed to Harold would commemorate a powerful aggressor, but one who eventually met their doom in a cataclysmic fashion in the field. Such an association might have been more powerful than any positive megalithic attribution to a legendary Welsh ruler.DSC03002

I’m not saying that we can believe these stones marked an assembly site, a battle site or a burial site of early medieval date. However, I do propose that sometime in the 12th/13th centuries, they may well have become linked to a legacy of Harold’s impact on the Welsh landscape, and the deeds of his enemies, as much as to Harold himself. These stones might have thus served in discourses of resistance to the Saxons and Normans with this attribution, as much as ‘colonial’ monuments, in a complex and fluid Anglo-Welsh borderland.

Heritage dimensions

There is a metal Ministry of Works-style sign, but actually it is far more recent, dating from the Cadw era (1986, i.e. post-1984). It is just decidedly ‘retro’. DSC02997

A more modern yet faded heritage board by the gate. It speaks of the archaeology and the legendary associations.


A further heritage aspect is that the gate is itself commemorative: a threshold enforcing association between stones and heritage organisations.



In summary, this is a fascinating site and well worth a visit. The potential of early medieval activity of some sort at this location – assembly, battle, burial site – and thus inspiring the Harold attribute, remains a temptation. Whether so or not, the Harold association does suggest a now-lost specific story linked to Harold Godwinson and circulating in the area in the Anglo-Norman period.


Whittle, E. 1992. A Guide to Ancient and Historic Wales: Glamorgan and Gwent. Cardiff: Cadw.


Offa’s Dyke south of Bronygarth

Re-posted and adapted from Howard’s Archaeodeath blog.

The fabulously preserved section of Offa’s Dyke south of Bronygarth can be found where the linear earthwork navigates the slopes on the south side of the Glyn Ceiriog.

IMG_2049IMG_2050IMG_2056I’ve recently posted in detail about this section here and I’ve also posted about it before here. I also mused on the presence of a boot hanging from the post of a stile.


The main reason for posting on this again here is to consider:

  1. a first-time visit with students: to show a picture of my students actually getting to see this (usually it’s a long walk to get to this location from the nearest easy parking, but we had our bus driver stop in the lane for a short while so we could complete this visit);
  2. the dyke’s scale and design: at this point it was possible to discuss with students the scale of the ditch in particular, where so often it isn’t preserved to this depth;
  3. the visual envelope of the dyke was considered too: to point out again how the dyke looks uphill and has a limited prospect in this location but amazing long-distant views back over (‘Mercian’) territory it was defending and even out over Cheshire towards Lancashire (‘Northumbria’);
  4. the dyke’s positioning: indicate how the monument descends and thus cuts across a major river-valley
  5. the dyke’s conservation at this point is good survival since it is incorporated into a field boundary, but we also noted the badger-damage to the earthwork;
  6. surrounding land-use: we observed how ploughed farmland encroaches right up to the back of the monument;
  7. the use of the linear earthwork for leisure activities was identified, as seen on other locations where the Offa’s Dyke Path navigates the back-side of the monument;
  8. present-day interactions from visitors were identified: graffiti on beech trees;
  9. aesthetics: we took note of the wonderful autumnal colours as the sun set on our successful field trip!




Exploring Offa’s Dyke at Craignant

Adapted from Howard’s Archaeodeath blog.

In October 2018, Dr Caroline Pudney and the final-year students explored sections of Offa’s Dyke. Having visited Montgomery Castle and then Dudston Fields, we headed north to explore a very different topographical situation for Offa’s Dyke.

The only similarity is that here, again, Offa’s Dyke follows the modern Anglo-Welsh border. As we headed north from Dudston Fields, we saw the dyke at key points including at Llanymynech where it heads for an Iron Age hillfort on a prominent hill having negotiated the Severn valley, and later on we saw it dramatically situated at Baker’s Hill and Carreg-y-big. We then walked a section of Offa’s Dyke from Craignant south up onto Selattyn Hill.

In contrast to the wide open vistas of the Vale of Montgomery, here Offa’s Dyke navigates steep topography as it jumps across and blocks tight valleys in the Welsh uplands.

Offa’s Dyke looking S as it descends around the western edge of Selattyn Hill

The line of the dyke was clear here, we could see how it bends eastwards as it descends into valleys, so its concave line more effectively overlooks and impresses those approaching from the west. This is a feature Ray and Bapty (2016) identify.

We then followed it as it skirted the western edge, rather than attempted to reach the top, of Selattyn Hill.

Looking south from Selattyn Hill over the Shropshire Plain, one can see the Wrekin and Long Mynd

We also discussed how you have fairly restricted views westwards from the dyke, contrasting with the situation at Dudston Fields. However, the significance of its line in navigating this more challenging terrain might have instead come from strategies of communication along the dyke’s line, rather than an ability to surveil extensive vistas west. Furthermore, as I’ve suggested before on Archaeodeath, the dyke in this area controls stupendous views eastwards, and the hills immediately behind it to the east might have served as beacons affording communication over 40 miles across the Cheshire and Shropshire plains.

The composition and design of the dyke differs here too. It is less monumental than at Dudston Fields. The dyke has to be dug into very different geology: here it is stone and must have been very difficult to excavate. The bank is smaller, and the ditch seems less broad, meaning it might have been deep but steep.

Another point of interest is that the rise up from Craignant reveals demonstrable scrapes that might be unrecorded primary quarries, used to excavate stone to face the dyke. It might have originally been dressed to appear like a drystone wall: a point suggested (again) by Ray and Bapty.

Traces of scoops that might be primary quarries used to construct the bank of Offa’s Dyke, looking W

An additional point of interest is the proximity of the dyke to a seemingly prehistoric cairn, although I’m not suggesting any precise connection to the dyke.

looking W over Offa’s Dyke, one can see a striking prehistoric cairn

The section at Craignant allowed us to identify the later commemoration of Offa’s Dyke: the notable Craignant tower. This is a 19th-century folly affording the dyke with a fortified Gothic appearance.

The Craignant Tower from the east heading W into Wales
Looking W from within Wales, the dyke rises steeply at an angle


Finally, we discussed the role of the Offa’s Dyke Path, and we met walkers enjoying it. The walk gave us a clear sense of the dyke as a long-distance footpath, and the signs revealed its intersection with the Shropshire Way.

Signs navigating the path
Offa’s Dyke path running along an old lane on Selattyn Hill


Dudston Fields, Offa’s Dyke

Adapted from Howard’s Archaeodeath blog.

Dr Caroline Pudney and myself led a field trip in October, and we were lucky to be joined by my new PhD student Liam Delaney who has recently begun doctoral research Offa’s Dyke’s landscape context. The undergraduate students got a treat of multiple locations and a sense of the entire northern part of the Anglo-Welsh borderland: its prehistoric hillforts, Roman sites, early medieval monuments and some of its castles too.


Having driven down the A483 from Chester, I used the microphone on the coach to point out the topography along the way and, specifically, where Offa’s Dyke and Wat’s Dyke ran. Then, when at Montgomery Castle, we looked out and identified where Offa’s Dyke navigated the Vale of Montgomery.


The first site where we visited Offa’s Dyke up close was at Dudston Fields in Chirbury parish, just east of the modern Anglo-Welsh border and not far east of Montgomery. I had visited this with the EMWARG conference delegates in 2015 for the first time, when Dr Keith Ray had expounded his arguments, subsequently published in his 2016 book about its adjusted-segmented design. This time it was great to have Liam there – his first visit – as well as to bring Ray and Bapty in book form!

IMG_1934On this visit, I got to see it more, less restricted by the EMWARG delegates. The students, Caroline and Liam were excellent for bouncing ideas off and adding key points about the monument and its landscape setting.

We started off by identifying the striking fact that this is the only section of Offa’s Dyke to be surveyed to a modern technical standard by the former RCAHME (now Historic England). I then pointed out and discussed the interpretation of the medieval ridge-and-furrow agriculture in relation to the dyke.

IMG_1937Next we explored what we think we know, and what we don’t know, about the dyke’s form and appearance. We also discussed Ray and Bapty’s arguments regarding the dyke’s ‘adjusted-segmented’ design of 20m-80m segments, each section aligned at slightly different angles. Here and elsewhere, as discussed for Hawthorn Hill, the dyke is made of a series of straight lengths, rather than being curvilinear. One student suggested this might have been to optimise the challenge of those scaling the monument and minimising erosion gullies.

IMG_1944We also talked with the students about the dyke’s differential survival depending on post-medieval landscape utilisation, the bank’s height and ditch’s depth vary considerably over this short stretch alone. In the field to the south, the monument has all but disappeared. In the Dudston Fields section, there are varying survival qualities of the bank and ditch in evidence.

We discussed how the monument is an ongoing victim of casual neglect. Fresh sheep-scrapes were in evidence into the bank. We also noted positive management strategies by the farmer: old logs dumped in the ditch to put off animals using sections where erosion was in evidence.IMG_1949

Then we turned to the broader landscape dynamics of the monument: its management of movement through the landscape, and surveillance of the landscape. We considered this in terms of both it managing the west-east traversing of the Vale, and perhaps also creating a protected corridor for traffic N-S. The idea that the dyke overtly ‘faced’ Montgomery, as proposed by Rap and Bapty, was also discussed.



In photographic terms, both students and sheep were very useful: affording a valuable scale to show the monumental dimensions of the dyke at this point.




Offa’s Dyke from Montgomery Castle

This is the first of a series of posts regarding an October 2018 full-day field trip with University of Chester archaeology students exploring the Anglo-Welsh borderlands, focusing on Offa’s Dyke. The aim is to get the third-year students on module HI6001 – Archaeology and Contemporary Society – out into the landscape to think about the history, archaeology, and heritage of borderlands and frontiers, and using the building and afterlife of Offa’s Dyke as a case study.

First up on our day was Montgomery Castle. I’ve blogged about this 13th-century castle before, and its ‘archaeodeath’ dimensions. As discussed on this earlier post, we explored the castle’s architecture, evidence of its attendant town wall, and also how death and memory features in its heritage interpretation.

IMG_1922I also used the castle to talk about the complexity and fluidity of the Anglo-Welsh borderland, and how castles had many functions in this regard. They were (in part) attempts to control the March of Wales militarily. Yet equally, together with their boroughs, they were attempts to socially engineer and exploit the borderlands economically and ideologically. We also discussed how the military significance of the borough and castle came to the fore once more in the early 17th century, in the English Civil War. The largest battle fought in Wales took place north of the castle. This revealed how, while not always a focus of conflict, the topography and locality had latent strategic importance down the ages.

We also discussed how the limitations of the heritage interpretation regime – with no phased plans or details of archaeological investigation allowed – provide only a partial appreciation of the castle.

I’d like to add one additional comment about its memorial dimensions. En route back to the bus, one of the students asked: why doesn’t this particular bench have a memorial plaque? I suggested an answer: its location is prominent – between the car park and the castle – but it lacked a long-distance view.  Perhaps no one wanted to be memorialised here for that reason. Its location is practical, but lacks the aesthetics of a panorama as afforded by benches elsewhere. So with the help of the students, I added something to my thinking about this site in memorial terms.

Anyway, the main point of visiting was to look east from it over the landscape and appraise Offa’s Dyke from this paranorama. To look east from Montgomery Castle is to view of the line of Offa’s Dyke as it traverses the Vale of Montgomery east of Montgomery and west of Chirbury and Church Stoke.  The photo above shows the famous Dudston Fields stretch from the inner ward of Montgomery Castle.

Ray and Bapty postulate that Montgomery Castle may overlay an earlier Welsh elite centre, so this close proximity might be significant for the 8th century AD. Sadly there is no definitive archaeological evidence in support of this. Moreover, there is no heritage interpretation pointing out the line and significance of Offa’s Dyke in an area that was to become a strategy later medieval frontier zone and where the Dyke follows the modern border for a significant length.