A hitherto unrecorded manuscript amulet with a Brittonic or P-Celtic inscription, circa AD 675 – AD 725


This blog aims to solicit feedback, comments and expert opinion on a remarkable hitherto undocumented Brittonic manuscript, with a view to eventually having a scholarly description and analysis published in an appropriate peer-reviewed journal. Suitably qualified potential authors of such an article are invited to contact the Site Administrator.

Formerly in the collection of Wallis Budge, the manuscript is a circular disk of scoured vellum, re-used from a very early Italian bible, with a 6 word inscription in a so far unidentified Brittonic or P-Celtic dialect.


The introductory notes given here regarding translation of the text and palaeographic identification of the locale and date should NOT be regarded as definitive, but simply as a provisional summation of the most widely prevailing critical views thus far. It’s very possible that they may be amended or altered entirely as a result of further research and analysis.

Likewise, there is absolutely no critical consensus or commitment thus far on which of the Brittonic or P-Celtic dialects is represented here.

The verso has a fragment of Genesis 6:12 in Roman Lapidary Capital script from an Italian bible, circa AD 425 – 450, likely one of the earliest surviving witnesses to the Latin Genesis text.

The recto is scoured, palimpsest, with an early Celtic cross in red ochre surrounded by the words “Deƿos me conaƿet – atir me snadit”. The most likely translation of this is  “God watch over me, Father protect me”.

On iconographic and palaeographic grounds the amuletic inscription dates from the late 7th or early 8th century and probably originates from Northumberland, likely the Monastery of Saint Peter at Monkwearmouth.

Written witnesses to any Brittonic  tongue longer than six consecutive words are generally limited to land charters of this (mid-to-late seventh to early-eighth century) period, and are nearly always confined to Old Welsh alone. None of the numerous experts consulted so far is aware of any comparable example.

That this is a palimpsest on an exceptionally early 5th century Latin Old Testament, in a fine Roman Lapidary hand, is remarkable in its own right.

Formerly in the collection of Wallis Budge, with his notes.

Just 6 words, but possibly one of the oldest surviving manuscripts in a native British language pre-dating the Anglo-Saxon era.

Dialect and locale

The oldest surviving Brittonic documents are written in Old Welsh (ieithoedd Brythonaidd):

Inscribed stones at Towyn (7th century)
The Surexit Memorandum added to the Lichfield (or St Chad) Gospels (8-9th century)
The Computus Fragment – Cambridge MS Add 4543 (10th century) Book of Llandaff – Llyfr Llandaff (12th century)

The primary Middle Welsh manuscripts are:
Black Book of Carmarthen – Peniarth MS I (12-13th century)
Book of Aneirin (circa AD 1250)
Book of Taliesin – Peniarth MS II (circa AD 1275)
Red Book of Hergest (circa AD 1375-1425)

The other Brittonic languages are attested from the 8th century (Breton) and from the 10th century (Cornish).

Old Welsh evolved from Brittonic (also called Brythonic), the Celtic language spoken by the ancient Britons. Alternatively classified as Insular Celtic or P-Celtic, it likely arrived in Britain during the Late Bronze Age or Early Iron Age and was probably spoken throughout the island south of the Firth of Forth. After the final withdrawal of the Roman administration, and the subsequent migration of Germanic speakers to Britain in the 5th century, the Brittonic language began to fragment, resulting in increased dialect differentiation, and evolving into Welsh and the other Brythonic languages (Breton, Cornish, and the extinct Cumbric). It is not clear when Welsh became distinct. Kenneth H. Jackson has suggested that the evolution in syllabic structure and sound pattern was complete by around 550. Old Welsh may have been spoken in both Wales and the Hen Ogledd (“Old North”), the Brythonic- speaking areas of what is now northern England and southern Scotland, and may therefore have been the ancestor of Cumbric as well as Welsh. The earliest Welsh poetry – that attributed to the Cynfeirdd or “Early Poets” – is generally considered to date to the Old Welsh period. However, much of this poetry was supposedly composed in the Hen Ogledd, raising further questions about the dating of the material and the language in which it was originally composed.

Some signs point to Cumbric as the Old Brythonic dialect of the amuletic text. However, the word “conaƿet,” used here, appears in its Old Welsh form (without the prefix) in the Llyfr Du Caerfyrddin (Black Book of Carmarthen, VII) as: “kyƿoethauc duƿ aƿet. y din iny deheu ƿuchet.” (Skene: “The mighty God will keep in his power the man of correct life”). Given what little we know of the written language of the Hen Ogledd, this may well be mimicked by Cumbric and others aside from Welsh.

The meaning and context here is:

“under the protection of” and that derivation is Anglo-Norman, from Old Saxon: “aƿet” (with the Latinate prefix “con” meaning “with the binding of,” here meaning to fall under the protection of God) derives from the Old Saxon apô.”

From this it follows that “aƿet” likely only reached Welsh after the 7th century, ie after the period in which this amulet was written. This would definitively exclude Old Welsh (and – less certain, but possibly – Cumbrian as well) from further consideration.

“The Latin prefix here wouldn’t have yet appeared in Welsh at this time owed to the relative insularity of the language (and likewise most probably Cumbric). The Common Tongue grew broadly from the Roman Conquest, and what we now have begun to call “Common Brittonic” would have gathered and acceded to many of the Latin forms after c. AD 450, but the more insular variants (or ‘dialects’) would have held out against them – perhaps for quite some time, as most scholars believe. Sadly, history records so little that modern P-Celtic scholars rely primarily upon place names, monument and grave inscriptions, and the handful of Charters listed in the Sawyer Catalogue whose Bounds were written in vernacular language(s), for their extrapolation of verb forms.”

Likewise, the text must fall under the Brittonic and not Goidelic families, since it contains one full word in transitional Anglo-Saxon – the permutation of “snade,” and hence leaves us with Cornish (yethow brythonek), Breton (yezhoù predenek), Pictish (to which there are no written witnesses), Cumbric (for which likewise no written witness exists), or Common Brittonic.

Cornish is unlikely – the syntax is unknown, and most of the language formed very late, having grown through and out of Anglo Saxon. (Gaulish can be rejected as well, as it seems to have died off circa 500 AD, and this inscription is too old for Early Scots).

Cumbrian remains a strong possibility:

“…the ‘snade’ you mention in Cyningsnade is a derivant meaning “public land” which would have been protected by the Crown (see Leo Heinrich’s Treatise on the Local Nomenclature of the Anglo- Saxons page 67) and there are no other comparators. In any common Brittonic tongue, this would have been a loanword from the new language (i.e. Anglo-Saxon, or Englisc as its native speakers called it) into the ‘old tongue’ which would conjecturally have become part of the written language towards the end of the seventh, or in the early years of the eighth century. Since Cyningsnade persists in Kentish and Northumbrian, I’d look to the learning centers – either of the two monastic houses in Canterbury for Kent, or in County Durham (Gateshead House, the Nunnery at Ebchester, Hartlepoole Convent, Wearmouth or Jarrow), although all were occupied by around the year 650, at least three fell or were abandoned early on (c. 800). South Shields House (Wherhale) can be safely struck from the list of origins as it was founded by Saxon monastics, despite the attribution to Aiden of Lindisfarne in modern times, and hence any inscription in the vernacular tongue there would have been in Old English.”

“Truly however, I believe you’re looking at an indigenous vernacular here, spoken (and written) by a transported people;; the cross decoration in red earth is a manifestation of the Irish monasteries (…) this red color was derived from bog iron, the Irish monks carried some of their precious pigments with them on the voyage to England. The language is almost certainly the product of north-central England. Much of the founding of houses on the eastern coast was carried out by Irish monastics crossing the Galloway Strait, skirting Burrow Head, landing at the mouth of Solway Firth, and then tramping cross country from the west to the east coast, beginning near the end of Ēadwine’s reign. The Cumbric tongue held fast in Northumbria, the people being slow to accept the New Tongue, and the Irish monks almost certainly drew with them Cumbraic people on their trek to Durham. Charles Phythian-Adams would agree;; see his paper (for the Antiquarian and Archaeological Society) From Peoples to Regional Societies: the problem of early medieval Cumbrian identities. His assertion that the people of Solway and the Eden Valley, right across Britain’s midsection, “continued to speak Cumbric throughout the Bernician occupation,” is well- supported, and would take us right through the time period when Durrow and Wearmouth were founded. Any local pilgrim setting out on the route followed by Willibald of Wessex would almost certainly have been a native Cumbric speaker, and the vernacular would have stretched well into modern day County Durham, sufficiently so that an amulet of this sort, written by a monastic house for a traveler, would have reflected the pilgrim’s language – here almost certainly being (off the record, not for publication) Cumbric.”

Breton has been suggested by one source:

“Benedict Biscop brought stonemasons from Francia to build the Monastery of Saint Peter in the early 670s and again, in 682 to build the second monastery at Jarrow. Breton must surely be amongst the top contenders as to dialect here.”

But rejected by another:

“While your theory that this was written for one of the Frankish workers to protect him on his journey back to Francia once his job was finished is solid, why wouldn’t the vernacular on the amulet have been a Franconian dialect rather than a Brythonic one?”

“There are still readable remnants of the lower script on the recto, though most were scoured away before the upper script was written, the vellum (parchment, more properly, as the quality of the skin is very poor indeed) having been re-used from an outdated book. I’d guess that the Monastery was in the habit of supplying blessings and portable scripture to pilgrims, as was often the case during the period, doubtless for a donation, of course, and probably in the vernacular of the region – here being Common Brittonic.”

“I would stop at ‘Brittonic’ and leave the dialect for others to conjecture, or at least read T. M. Charles-Edwards’ studies regarding the faint evidence, and the huge similarities of Cumbric with Cornish and Breton.”


The Monastery of Saint Peter at Monkwearmouth

The Monastery of Saint Peter at Monkwearmouth was founded in 674 by Benedict Biscop, with the establishment of the monastery  at Monkwearmouth on land given by Egfrid, King of Northumbria. His idea was to build a model monastery for England, sharing his knowledge of the experience of the Roman traditions in an area previously more influenced by Celtic Christianity stemming from missionaries of Melrose and Iona. A papal letter in 678 exempted the monastery from external control, and in 682 the king was so delighted at the success of St Peter’s, he gave Benedict more land in Jarrow and urged him to build a second monastery. Benedict erected a sister foundation (St Paul) at Jarrow, appointing Ceolfrith as its superior, who left Monkwearmouth with 20 monks (including his protégé the young Bede) to start the foundation in Jarrow.

Benedict brought workmen from Francia to build these churches, the first ecclesiastical structures in Britain built of stone, and furnished it with glass windows, pictures, service books and the library he had collected on his travels. Window glass being unusual in England at the time, Benedict imported glassmakers from Francia, who established a workshop at the Monkwearmouth site, which stands on a nearby site on the river Wear.

The two monasteries were so closely connected in their early history that they are often spoken of figuratively as one, Jarrow, despite being seven miles apart. Benedict himself was the first abbot, and the monastery flourished under him and his successors Eosterwine, Ceolfrith, and others, for two hundred years. Benedict, on leaving England for Rome in 686 established Ceolfrith as Abbot in Jarrow and Eosterwine at Monkwearmouth but, before his death, stipulated that the two sites should function as ‘one monastery in two places’.

Ceolfrith as abbot continued Benedict’s work in establishing the monastery as a centre of learning, scholarship, and especially book production, during which time a distinctive house style of half-uncial script emerged. Ceolfrith’s major project was the production of three great “pandect” Bibles (i.e. manuscripts containing the entire text of the Bible), intended to furnish the churches of St Peter’s and St Paul’s, with the third copy earmarked as a gift to the Pope. The only survivor of the three bibles is the Codex Amiatinus, now in Florence, the oldest complete surviving Bible in the world, which was being carried to Rome by Ceolfrith himself when he died in 716. The Bible was carried to Rome by his companions, whence Pope Gregory II sent his thanks to Ceolfirth’s successor, Abbot Hwaetberht.

The library Benedict had created on his travels to Rome and then given to the monastery made it the cradle not only of English art but of English literature – Jarrow is where the Venerable Bede received his early education under Ceolfrith’s patronage and lived, wrote and died as a monk. By his death Bede had established himself as England’s leading scriptural and historical authority, and was to have a vital post-mortem influence on the fortunes of the monastery. Bede’s writings, most importantly his Historia ecclesiastica gentis Anglorum, became so popular in the 8th century that they not only assured the reputation of the houses, but influenced the development of Monkwearmouth-Jarrow’s distinctive insular minuscule script, developed to increase the speed of book production.

The golden age of Monkwearmouth-Jarrow began to draw to a close in the late 8th century, as Northumbrian monasteries became vulnerable to Viking raids, with Monkwearmouth- Jarrow itself being attacked in 794 (the second target in England of the Vikings, after raids on Lindisfarne in 793). They were finally destroyed by the Danes about 860, and seem to have been finally abandoned in the late 9th century.


Several scholars were consulted on this specific issue.

The following comments, assembled from several different emails, are quoted verbatim, but sans attribution of the authors:


“I’ve again looked closely at the amulet – specifically at the edges – and it’s clear that the lower script wasn’t “added.” Find your fountain pen, cut a circle from a piece of paper, and try writing from the very edge of the circle across to the furthest edge. I’ve just tried it. The nib catches against the edge at the beginning, and flops over the edge at the end – in both cases leaving little blobs of ink at the edge. It’s clear that the circle was cut after the lower text had been written.”


“Could the lower text have been written across the vellum and the circle then cut through? Yes – but setting the text down – ruling in precisely the correct fashion – including the vertical rule at the left margin of the verso, or the right margin of the recto, and showing the single prick-mark in the vertical rule, just above and to the left of the “c” in “corruptam” – precisely as it should appear – then doing the mathematical calculation for both the length of line and the number of lines and the text itself to match with the recto’s now very faint “est terra iniquitate a facie eorum et ego disperdam eos cum terra…” – and then scouring that text away convincingly so that only a ghost remained over which the upper text would be written – would be an incredibly difficult undertaking, and the layout alone – even for a savant whose abilities extended to being able to “visualise” the lines, spaces, shortened lines owed to the use of the sigla and abbreviated Nomine Sacra, and verse incipits – would be challenging even had the entire “leaf” of the lower text been written both recto and verso to ensure continuity.

Then there’s the matter of the text itself. Whole text comparitors between the Vetus Italica and the Vulgata were only just being undertaken in the 1920s; this particular passage regarding The Flood is examined textually in the Appendices of the Yale Studies in Volume 63; page 221 – 222. Imagine [a hypothetical forger], going to the immense trouble of creating a double- sided facsimile in this fashion, then adding the upper text in two additional wholly different inks, using a specific monastic hand – with its curious tells [….]”


“For the sake of argument, let’s give [the hypothetical forger] savant status, and presume that he’d written the leaf out en toto, sigla and all, on foolscap first so as to determine where each line would fall, and then reproduced – in three exceeding different inks – the portion required (the upper margin and three lines) for the “base” of the amulet, first ruling the old vellum he employed to create this exactly as it would have appeared a millennium earlier, and then writing out the text at both sides, cutting it, scouring the recto so as to leave a ghost of the text behind before overwriting the upper text, including the cross, copied from a manuscript which he was [very unlikely to] have seen, and then – without allowing it to set for a thousand years so the ink of the lower text could adequately take hold in the vellum’s fibres – laying out the upper text in a circular manner, in a language which only he and a few other living men could have understood.


“Could anyone, no matter how well-educated, have amassed the various skill sets to piece together the language forms needed, to grind and prepare the various inks (and clearly, the inks match with their respective periods very precisely), to procure the exceeding early blank vellum, to lay out the whole of the lower script, wear it away without taking the vellum’s colour and patina (this is clearly impossible to do with abrasives such as fine glass-paper, pumice or rottenstone as they’d wear away the surrounding layers), and managed to create these three “things” – as they are three distinct “things” from whole cloth, sufficiently well to have fooled Budge and his colleagues? ”


“Agreed. Even today, with modern scientific tools, this would be a feat of the highest magnitude to replicate, and would require the cooperation of many different experts – and knowledge of the cross in the upper level would be, as you’ve noted, beyond his ken… but then I was convinced simply in having handled it.”

Lorica Invocations

The identity of the amuletic prayer has not yet been determined, but it has similarities to some early Celtic loricas.

In the Christian monastic tradition, a lorica is a prayer recited for protection. The Latin word lorica originally meant “armor” or “breastplate.” Both meanings come together in the practice of placing verbal inscriptions on the shields or armorial trappings of knights, who might recite them before going into battle.

The idea underlying the name is probably derived from Ephesians 6:14, where the Apostle bids his readers stand, “having put on the breast-plate of righteousness”.

Notable loricas include Rob tu mo bhoile, a Comdi cride, which in its English translation provides the text for the hymn Be Thou My Vision, the Lorica of Laidcenn and the Lorica of Saint Patrick, which dates back to – at least –  the 8th century.

An extract from St Patrick’s Lorica in both the original Gaelic and in English translation: