The 5th Century Chedworth Mosaic

It is evening. The sun casts long shadows and the man lingers a moment beside the shrine, watching the life-spring of his home trickle into the octagonal basin. He turns and walks the length of the corridor, up the stone steps, along the passage and finds his wife in the dining room.

This image has an empty alt attribute; its file name is capture1.jpg
The octagonal basin within the Nymphaeum shrine at Chedworth Roman Villa, Gloucestershire

‘Did you hear the news ?

‘Yes…we are governed by selfish incompetents. How on earth will we be able to manage here in the future?

‘I think we have enough for now…but it’s the children I worry about.’

By the end of the 4th century, Chedworth Roman Villa was at its best. A fine home for a wealthy family.

Whatever became of it? How did such a place become a ruin?

Archaeologists investigating the Romans generally depend on an abundance of finds for dating. Coins, pottery and all the other lovely things that the Empire enabled merchants to import from around the Roman world.

This world gradually fell apart. The tap was turned towards off in the 5th century. Some coins enter Britain in the early 400s and there was some pottery production. A few shipments of exotic wine made it as far as Gloucestershire.

There is a piece from a 6th century Palestinian amphora unearthed at Chedworth. Could there still be people living at Chedworth able to afford such things?

Whatever…. finds are few and generally the events of the 5th-7th centuries are tough for archaeologists to unravel.

A challenge then: particularly as the upper archaeological levels were stripped away and discarded in 1864, when the Villa was discovered, and then rapidly excavated down to its mosaics.

So…come and visit Room 27 with me in Chedworth’s North Range. It is 2017 and I am excavating the trench in the north-east corner.

I must warn you….I am going to talk stratigraphy at you… I have to I’m afraid, you won’t believe me otherwise. Archaeologists live and breath stratigraphy. How can things be proved without it?

So, I am kneeling in my trench: In front of me is the north wall of the room and immediately to my right, beyond the east wall, are a line of archaeologists in Room 28.

This image has an empty alt attribute; its file name is capture-2.jpg
Looking across the wall from Room 27 watching the Room 28 mosaic being uncovered.

They are working backwards from the north wall carefully uncovering and cleaning a mosaic. They are finding a row of circles containing three and four petalled flowers alternating with woven knots and linked by woven strands of guilloche in red, white and blue tesserae.

Their room is more exciting than mine. I only have a thin band of a plain crushed tile and mortar floor (opus signinum) surviving against the north wall. Below this is the floor’s mortar bedding and below this the floor’s mortared limestone hardcore which survives across my trench.

This image has an empty alt attribute; its file name is capture-4.jpg
Looking east within Room 27 the surviving piece of opus signinum floor is against the north wall on the left The hardcore it was built on is to the right of the ranging pole with its 0.2m long red and white divisions. I have taken out a section of it to reveal the dark soil against the east wall at the top of the picture. This east wall runs up to the north wall but is not bonded to it. A later insert.

I told you I’d talk stratigraphy. This is the important sequence of events, most recent at the top and the earliest at the bottom…. and what you extract from each of the distinctive layers is important to enable you to unpick the past…. century by century.

You know…. if you went on holiday and forgot to cancel the papers… they’d pile up below the letter box. The earliest would be at the bottom.

So.. the hardcore which supported the crushed clay tile and mortar floor. Well, it covered the foundation trench for the wall between Rooms 27 and 28.

We have come to the point in the sequence of events when the wall was constructed.

The builders dug the trench, placed the foundation stones for the wall in the trench and then shovelled soil and rubble..and anything lying about… back into the trench, packing it against the newly built wall.

This image has an empty alt attribute; its file name is capture-5.jpg
Looking down on my trench. The inserted east wall on the right and the north wall it was built against it at the top of the picture. At this stage, I have left the strip of opus signinum floor against the north wall on its thin bedding layer of mortar. This lay above a gravel and mortar hardcore layer about 5cm thick and this covered the dark loamy soil filling the foundation trench. In this soil I found a fragment of black pottery some charcoal twig fragments and two small fragements of animal bone. This foundation trench cut through the creamy yellow limestone fragments set in clay which was the natural bedrock. This can be seen on the left hand side of the trench. When I excavated a small section against the north wall, it could be seen that the foundation trench of the east wall also cut the foundation trench of the north wall.

The technical archaeological term for this is ‘the foundation trench filling’ and anything found in this helps date the construction of the wall. My boring opus signinum floor and my neighbours’ exciting mosaic floor must be later than the wall because my floor is built over the foundation trench filling. You cannot lay out a mosaic design to fit a room until the walls are built. That makes sense doesn’t it? Hold onto that thought.

In August 2017, I looked carefully for finds from the soil of this precious ‘filling’, a coin would be excellent in such a context…but no, all I got were fine strands of charcoal twig, two small fragments of animal bone (traces from a meal I suppose) and a single black piece of pottery.

This image has an empty alt attribute; its file name is capture-7.jpg
Pete’s trench in the south-east corner of Room 27. The walls need deeper foundations here because the natural valley slope, that Chedworth is terraced into, drops away to the south and to create a level floor surface material needs to be brought in. Notice the completely different designs of the wall footings. On the right hand side of the picture; the south wall, like the north, is of regular courses of nicely faced stone; whereas the east wall, on the left hand side, has a cap of roughly dressed stone on top of a heap of rubble with bits of tile in . It is clear that this is later than the south wall because it is built as a straight joint against it.

This east wall foundation had cut the foundation trench of the north wall and the stones of the east wall abutted the north wall…. What I’m saying is that the east wall was not part of the original construction of the North Range.

Pete had dug another trench at the other end of Room 27 in the south-east corner. The soil was much deeper there. The building had been constructed into a valley slope. It was cut into the bedrock on the north but the foundations needed to be much deeper to the south and to make a floor, lots of soil needed to be brought in to create a level surface. A wedge of soil above the sloping bedrock deepest against the south wall.

The style of construction of the south wall looked much the same as the north wall and the east wall butted up against it and was clearly a later construction. There were sherds of pottery and charcoal in the foundation trench of the south wall….

So the contemporary north and south walls were continuous through the space which became Rooms 27 & 28. When our new wall was inserted it became the east wall of 27.. which of course…was the west wall of 28.

This image has an empty alt attribute; its file name is capture-6-1.jpg
Drone photo (copyright Mike Calnan) of the east edge of Room 27 and mosaic room 28. The mosaic pattern is lost in the centre and in the bottom left quadrant of the room are the remains of two later hearths with traces of burning around them. Notice how the surviving mosaic runs up against the wall top left and how the whole mosaic pattern has been made to fit this room. Beyond wall on the far left, is my trench (top left) in the north-east corner of Room 27 (just above the rolled up white geotextile matting). This is where the radiocarbon dates were taken..from the foundation trench on the left side of the dividing wall. Pete’s south-east Room 27 trench is bottom left on this picture.

We soon found out that the mosaic in Room 28 had been worn away in the centre of the room. There had been a workshop here. Two fireplaces or hearths had been made out of reused bits of villa and built into the burnt eroded centre of the room.

At the end of the 2017 excavation, we thought we had the answer. In the 4th century, a new wall had been built in the North Range to create two new rooms. A plain floor was constructed in 27 and a new mosaic created for 28. By the 5th-6th century, Chedworth was falling apart. The economy had crashed and the once rich owners had abandoned Its beautiful mosaic rooms…..it was not being looked after.

Instead, 28 had been turned into a workshop…..or so we thought.

This image has an empty alt attribute; its file name is capture-10.jpg
Looking again at the finds that Nancy and her team had recorded and catalogued from the 2017 excavation at a picnic table within the mansion house courtyard.

Join me now at a wooden picnic table in the summer sunshine of 2019. We are in the stable-yard of a National Trust mansion house where Nancy and her volunteers have finished processing the Chedworth finds.

The charcoal strands from the foundation trenches of 27’s south and east walls, along with the ash from 28’s late hearths will be sent for radiocarbon dating. The pottery from the south wall foundation trench looks good for the 2nd century. A flanged bowl with acute cross-hatch decoration is particularly appropriate. Jane the pottery specialist will check it out…I wonder what she will make of my black bit from the east wall trench filling.

This image has an empty alt attribute; its file name is capture-11.jpg
Pottery finds from the Room 27’s south wall foundation trench filling. The sherd bottom right is part of a ‘flanged’ bowl which comes from the kilns around Poole Harbour in Dorset. The flange is flat and angled from the rim of the bowl. the cross-hatch decoration inscribed on the side is acutely angled and decorates a broad area of the side of the bowl…. I tell you this because it is typical of 2nd century Black Burnished pottery produced in Dorset at that time. It backs up the radiocarbon date.

The radiocarbon dates come back first. The charcoal from the south wall matches the pottery …mid to late 2nd century. We have had similar dates from other parts of the early North Range.

Then I see the date from the charcoal found in the foundation trench of the inserted east wall…AD424-544 at 95.4% probability !! That’s not even tentatively into the 5th century…it might even be 6th. I contact the mosaic specialist.

‘Have you come across any British 5th century mosaics?’

‘No, the economy collapsed, coinage and pottery production disappeared. Would a mosaic business survive? Would a villa have the confidence and wealth to redesign the house and lay new floors? Anyway… what about the 5th – 6th century workshops in Room 28’

‘The radiocarbon dates say they’re 12th to 14th century. Medieval rather than Dark Age’

‘The dates must be wrong. The Room 28 mosaic is one of the later more poorly constructed mosaics. There are lots of mistakes in the design but I would need a lot more proof before I could believe that it was made in the 5th century.’

I ask around.There are hints.of late mosaic floors but radiocarbon dating within British Roman villas has not been common.

I need confirmation. Nancy gets the larger of the two pieces of animal bone found in the foundation trench and Mark, Chedworth’s manager, agrees the funding to send it for a second radiocarbon date. It could be just a stray piece knocking around the site from an earlier period…but it’s worth a try. The result will take several months to process.

Then the pottery report comes back from the specialist. The fragment of black pottery from the trench turns out to be Late Roman Shelly Ware. It dates from after AD 360…it could be much later but nobody knows when production stopped for this ceramic type…anyway it confirms that our inserted wall was at least a latest Roman construction.

We waited…and waited… and eventually our second radiocarbon date was ready….

Not as clear cut as the charcoal date but the bone date definitely supports it. In the 95.4% probability band the date is split AD 337- AD432 (87%) and AD491 – AD531 (8.4%).

The radiocarbon date is measured from when the animal died or the wood was cut and burnt. It then becomes debris to lie around and then fall into the building trench for a new wall…..against which the mosaic floor was later built.

We have at least a 5th century mosaic at Chedworth…could even be 6th century but that would be pushing things a bit. There are two other mosaics in the North Range with Room 28’s late style. The corridor mosaic (Room 33) is particularly quirky.

This image has an empty alt attribute; its file name is capture-8.jpg
The mosaic of the corridor of the North Range. A similar level of skill to the Room 28 mosaic. This is phase 3 of mosaic design at Chedworth…OK but full of errors and mosaic making past its 4th century best. This one could also be 5th century but there is no surviving evidence to date it.

Anyway, time to go back to the worried owners of Chedworth Roman Villa…. having their conversation in the dining room…one evening at the end of the 4th century.

‘Don’t worry. The kids will be alright. Cirencester (Corinium) and the rich villas surrounding Roman Britain’s second largest town will keep the Romanised flame flickering for a little while after the Empire’s soldiers sail away’ …

but……perhaps the great great grandchildren should watch out.

In a very rare historical survival from the early 6th century, the British monk Gildas writes of corrupt government and warns of trouble brewing …but according to the Anglo Saxon Chronicle, the Saxons didn’t defeat the Romanised British kings of Bath, Gloucester and Cirencester until the Battle of Dyrham in AD 577.

Plenty of time for new 5th century floors to be made in West Country villas. Let’s find some more.

12 thoughts on “The 5th Century Chedworth Mosaic

  1. Now that is interesting… Is there evidence for the individual pieces being re-used or maybe the whole mosaic being relaid, maybe moved from elsewhere?

    • Dear Martin Good points to consider but the mosaic fits the space nicely, an intricate design but has some mistakes in the layout. No evidence that it was moved from somewhere else.
      With best wishes Martin

      • Thanks Martin. I was wondering about whether this is evidence for late manufacture or architectural salvage, Given the different craft skills necessary for each of these, it is particularly interesting. Thanks for sharing.

  2. Absolutely fascinating. Given the quirky workmanship of some mosaics elsewhere, coupled with sophisticated imagery, it wouldn’t surprise me if others were later than we have tended to think. Definitely something to bear in mind.

    • Dear Pat

      Good to hear from you. The C14 dates give us the freedom to expand the date range of mosaics in Britain. Previously, Chedworth’s 3 mosaic phases had to be placed in a tighter dating framework though the styles differ considerably as you know.

      With best wishes

      Martin

  3. What an enjoyable read in these times of COVID. So good to hear from you and always interested in hearing about the work you and your team do.

  4. Fantastic confirmation Martin of what you hoped-congratulations!
    Super clear explanation of what you/we found. I feel so honoured you include a picture that I feature in; that time was SO exciting to me. Thank you again for allowing me to be involved.
    Best regards, Helen

  5. Martin; fantastic stuff, and beautifully written as ever. If the workshops in Room 28 are C12th-C14th does that imply the building was still standing then? Is this a surprise? The reuse of older sites always fascinates me – did our ancestors ponder on who had gone before them in the way that we do?
    Martin (yet another one)

    • Dear Martin

      It was so good to meet you in Bath all that time ago in March. What you say seems very likely though difficult to prove. There is a great Saxon poem written of the city of Bath …I remember a line but I can’t give you the source. ‘Strange to behold are the stones of this wall, broken by fate’ With very best wishes as ever Treasurer’s House work a long time ago now Martin

      • It was great to see you Martin. The poem is ‘The Ruin’, in the Exeter book. I did a dig just outside Exeter a few years back where in the C7th-C8th there had been crafting activity within an enclosure originally constructed in the Bronze Age and still surviving then as an earthwork. Some of the crafting involved pits dug into the sandy soil which had been shored up with a variety of materials, including ashlar blocks of both volcanic rock from the Exeter Series and limestone from Beer. Presumably a substantial Roman building had been robbed for this material – either from the city itself or a now unknown structure nearby. I find this reuse and continuity fascinating. Should get published in the DAS proceedings in a year or two.

  6. Dear Martin,
    Did the excavation in room 28 show that the mosaic floor had also clearly sealed the foundation trench on that side of the wall and did the foundation trench extend underneath the decorative part of the mosaic or just under the plain border? That would prove that the wall dividing 27 and 28 could not just be a rebuild of an earlier, less substantial division, with the op sig floor added afterwards.
    Cheers,
    Richard

    • Dear Richard That’s an interesting thought. The mosaic has the same border all round not just on the west side and it is integral with the mosaic design. There could have been a timber screen there which was replaced with a substantial structural wall with a deep foundation in the 5th century on the exact alignment of the original screen along its east face.

      We did not lift the mosaic to investigate whether the foundation trench existed on the east side of the wall in Room 28 as well as on the west side in Room 27. At the time, we had no expectation that the wall was later than the 4th century.

      I wondered whether the 5th century date might relate to a blocked doorway perhaps rather than the construction of the whole wall but the north end of the Room 28 mosaic is well preserved and there is no trace of wear on the mosaic to indicate a door ever existed there, neither is there any evidence of a straight joint or difference in build at this north end of the wall compared with the south.

      Dividing a space with a thin screen and then creating a substantial wall taking account of the sloping bedrock towards the south seems to indicate that the wall was structurally necessary but it might be argued that later alterations to the villa required a more substantial wall. Perhaps another floor being built on the villa which required a stronger wall to support it.

      I do plan to request SMC for collecting further samples including from the east side of the wall where the mosaic has been lost as the dating evidence is so interesting that it requires further sampling.

      With best wishes

      Martin

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s