Published: 09/03/2026
Updated: 10/03/2026
The rain was driving sideways across the churchyard at St Cadoc’s church, Llangatwg Feibion Afel (Llangattock-Vibon-Avel) when we arrived that grey afternoon. It was the kind of Welsh weather that seeps through supposedly waterproof jackets and turns gravel paths into streams. Our boots were already caked in mud from tramping around the graveyard, squinting at worn inscriptions on stones tilted at drunken angles after centuries of Welsh rain and wind.
We'd almost given up for the day when we ducked into the church porch for shelter, armed with torches to bolster the dying midwinter sun.
Propped against the east wall was a freestone slab, its surface carved with a distinctive post-Reformation crosshead common to this corner of Monmouthshire. But it was the inscription below that made us pause, rainwater still dripping from our coats. The text was bilingual (English facts followed by Welsh poetry) and there was something peculiar about the lettering. In the Welsh section, the capital Ws appeared doubled, and twice we spotted backwards Ns staring at us like mirror images.
Why only in the Welsh? The English text above was perfectly conventional. We photographed it carefully, uncertain yet that we'd stumbled onto something significant.
The inscription read:
HEARE VNDER LYETH
THE BODY OF JAMES
WATER DECEASED
THE 18 DAY OF APRIL
ANNO DOMINI 1696
Followed by the Welsh verse with its strange lettering:
GWEDDIWN BAWB AR
Y IESV HWN DDYCH
ON YN HAWDD NHEL
PV A DANGOS YN NHY
GWIR OLEVNI PWY FA
DDE Y GWIRION YN Y
GWELY
The mystery deepens
Back in the archive, we began the detective work. John Hobson Matthews had been here before us in May 1903. In a letter to Archaeologia Cambrensis, he described finding "a Welsh inscription on a freestone slab lying flat on the ground" in the churchyard, "immediately east of the south porch."
Four years later, Joseph Bradney saw it too, recording it in his History of Monmouthshire as a "flat stone" near the south porch.
But we'd found it inside the church, propped against the porch wall.
When had it moved? More intriguingly, who had moved it, and why?
The trail went cold. In September and October 1988, the Gwent Family History Society surveyed the church and published their findings on microfiche. By then, the stone was listed as being in the "Interior of the church" and we can assume they meant the porch. They transcribed the English inscription accurately, but made no mention of the Welsh poem (Though this was not unusual for them; they often noted the presence of Welsh poetry without transcribing it.)
Somewhere between 1907 and 1988, James Water's gravestone had vanished from the churchyard floor and reappeared inside. No church records documented the move. No one we contacted could remember when or why it happened.
F. G. Levett's 1984 survey of the church's commemorative stones didn't mention it at all, he simply quoted Matthews' 1903 description of it lying outside. We suspect he never actually saw the stone himself. E. M. Evans' 1997 Cadw report listed it among the porch stones, but summarised several with the phrase "As Bradney," suggesting she might have misread Bradney's 1907 account as placing it inside rather than out.
The gravestone had performed a disappearing act, and no one had noticed.
Following the clues
We turned our attention to decoding what the stone actually said. The bilingual inscription itself held clues to a vanishing world.
Why English for the facts, Welsh for the poetry? To modern eyes, it seems an odd choice. Whoever commissioned this stone was clearly fluent in both languages. Why not use one throughout?
The answer lies in the mid-16th century, when Wales was formally incorporated into the English state. English became the exclusive language of all official business. A gravestone, in one sense, functioned as a legal record of death, making English the appropriate choice for names, dates, and facts.
But Welsh remained the language of the soul. It was still the language of religious observance, of prayer and contemplation. When the time came to express grief, hope, or piety, Welsh was the natural choice. This pattern of linguistic division persisted across Wales for centuries, appearing on gravestones well into the 20th century.
The Welsh poem itself was crammed awkwardly onto the stone, forced into too narrow a space. When we reconstructed it properly, we could see it was a four-line verse with all four lines rhyming:
GWEDDIWN BAWB AR Y IESV
HWN DDYCHON YN HAWDD N HELPV
A DANGOS YN NHY GWIR OLEVNI
PWY FADDE Y GWIRION YN Y GWELY
W M
[Let us all pray to Jesus / who can easily help us / and show in the house of true light / who would forgive the innocent in their bed.]
The stone-carver had broken words mid-syllable—DDYCH/ON, NHEL/PV, FA/DDE—to make them fit. You can almost see him realising too late that he'd miscalculated the space.
The poem itself is remarkable. Welsh gravestone poetry from 1696 is exceptionally rare, and this particular verse appears nowhere else in the historical record. We've found no variants, no other uses. The identity of "W M," the poet, remains unknown.
But the poem itself revealed secrets. The dialect gave the poet away as a southerner. Look at the rhyme scheme: IESV, HELPV, OLEVNI, GWELY. In south Wales, all four words end with a high front vowel "i" sound, they rhyme perfectly naturally. But a northern Welsh speaker would never have constructed this verse, because in northern dialect, only OLEVNI has that "i" sound. The other three—IESV, HELPV, GWELY—would use the distinctive northern "u" (yr "u" ogleddol), a high central vowel. They'd rhyme with each other but not with OLEVNI.
There's another dialectal clue: FADDE. In standard literary Welsh, this would be FADDAI, with a diphthong. But in spoken Welsh, particularly in the south, people simplified it to FADDE; this is informal, colloquial speech rendered in stone. The later southeastern change to FADDA wouldn't emerge until well after 1696.
This was a local poem, written by someone who spoke the Welsh of this corner of Monmouthshire, who knew James Water, and who composed a verse that followed rhythmic and rhyming conventions but deliberately avoided the strict metres of cynghanedd poetry.
The parish records gave us our first breakthrough. James Water, it turned out, was also known as James Walter; a small but significant detail. The burial register for Llangatwg Feibion Afel recorded: "... son of James Walter and Eliza Walter was buryd 21 April 96."
The dates matched. The gravestone said he died on 18 April 1696; the burial three days later was entirely consistent.
But it was his probate inventory, drawn up on 1 May 1696, just two weeks after his death, that revealed the puzzle at the heart of James Walter's life.
The inventory began modestly enough. His "wearing apparel" and household possessions were listed: "one flock bed and bolster and its appurtenances," "brass pewter and household stuffe." The total value: just over £20. He was a tenant farmer (there's no mention of land ownership) but he owned his own "implements of husbandry" and the animals on his farm.
The timing of his death in spring shows in the detail: "a yoake of oxen," two cows and two yearlings, a mare, pigs, sheep, and lambs. Those lambs were particularly important in April, the season of new life on a farm. He had £5 worth of corn in store. The yoke of oxen confirmed he was growing crops as well as keeping stock. The farm-related possessions totaled a little over £22.
A modest estate, then. A tenant farmer of limited means.
But then came the shock: £180 was owed to him "by billes, bonds and other writeings."
To put this in perspective, the debts owed to James Walter were more than four times the value of everything else he owned. This wasn't a small loan to a neighbour. This was serious money-lending. The details weren't specified. We don't know who these debtors were or why a tenant farmer was extending credit on this scale.
The questions multiply: Where did James Walter get £180 to lend in the first place? What business was he in beyond farming? Was he acting as an informal banker for his community? How did a man with a modest farmhouse and a yoke of oxen come to be owed sums that dwarfed his visible wealth?
His bond mentions a brother, Thomas Walter, and identifies James as a yeoman, a status that implied some standing in the community but not wealth on this scale. The inventory reads like two different lives: the visible life of a small farmer, and the hidden life of a creditor handling substantial sums of money.
We still don't have answers.
This gravestone matters for reasons beyond James Walter's mysterious finances. It's a crucial piece of evidence in understanding the death of a language.
Welsh place names blanket Monmouthshire, testimony that Welsh was once the natural tongue of this landscape. The county boundary was not a language boundary, either. Welsh was spoken in Herefordshire well into the eighteenth century. But by the time census returns began recording language use from 1891 onward, Welsh had almost completely vanished from areas east of the Usk valley. Anglican church records show the shift to English services began as early as the 18th century.
James Walter's bilingual gravestone from 1696 proves Welsh was still alive and in active use here at the end of the 17th century, not just as a relic, but as the language of poetry, prayer, and intimate expression. The dialect features confirm this wasn't imported from Welsh-speaking heartlands; it was local, rooted, everyday speech.
This makes the stone a temporal landmark. It stands near the beginning of a long decline, a witness to a moment when Welsh still held its place in this community but English was already asserting its official dominance. The linguistic split on the gravestone, English for the state, Welsh for the soul, captures a culture in transition.
Much work remains to understand precisely how and when Welsh disappeared from eastern Monmouthshire, but James Walter's stone gives us a fixed point: in 1696, in Llangatwg Feibion Afel, Welsh poetry was still being carved into stone, still rhyming with the authentic sounds of local speech.
We know more now than we did on that wet afternoon in that church porch. James Walter was a tenant farmer of modest means who somehow came to handle substantial sums of money. He was buried in April 1696, mourned in a Welsh poem by an unidentified poet who signed himself "W M." His gravestone was carved with strange backwards Ns and doubled Ws that still defy explanation, and the verse was squeezed awkwardly onto stone by a mason who miscalculated his space.
The stone lay flat in the churchyard for at least 200 years, noted by antiquarians in 1903 and 1907. Then, sometime in the eight decades that followed, someone moved it inside to the shelter of the porch.
But who? And when, exactly? And why?
The church records are silent. No one we've contacted remembers. The gravestone simply migrated from ground to wall, and no one thought to document why.
Perhaps it was moved for preservation, to protect the inscription from further weathering. Perhaps there was a churchyard reorganisation and this stone, with its unusual bilingual text, seemed worth saving. Perhaps someone recognised its historical importance.
We'll likely never know.
What we do know is this: a stone that began as a memorial to one man has become something larger, a window into a moment when two languages still coexisted in one community, when tenant farmers could somehow hold substantial debts, when Welsh poetry was still being composed in this corner of Wales.
James Walter's gravestone moved once in the past century, slipping from its original position to its current home with no record of the journey. But the words carved into it have stayed fixed, still speaking across more than three centuries, still asking questions we're only beginning to answer.
Friends of Friendless Churches and St Cadoc's church
Facing closure since 2018, St Cadoc's church, Llangatwg Feibion Afel (Llangattock-Vibon-Avel), was rescued by Friends of Friendless Churches, joining our collection of historic places of worship in March 2023.
Back then, St Cadoc's church needed a lot of repairs – to the roofs, drains and the beautiful but very soft Old Red Sandstone. Thanks to the Cadw grant and the generosity of our supporters, urgent repairs to the fabric of the church were carried out in 2023, including conserving the glazing to the east and west windows and the Rolls Chapel.
This year, the Friends will carry out more repair work, due to commence in summer 2026. Thanks to the generosity of our supporters, we will be able to preserve the fabric of the church's tower and improve the internal heating arrangement. For now, St Cadoc's church, Llangatwg Feibion Afel (Llangattock-Vibon-Avel), is open daily for everyone to enjoy, but please check our website for the latest update on opening.
If you would like to support our work caring for 71 historic places of worship in our care like St Cadoc's, please consider donating or joining us as a member. Thank you!
Gwen Awbery
Madeleine Gray
(A fuller version of this blog post, with all the academic stuff like references, will appear in 2026 in volume 41 of the Monmouthshire Antiquary)
