James of St George: the master mason behind Caernarfon Castle

Who he was

Master James of St George was a master mason — in modern terms something closer to a chief architect and site engineer combined — who came into Edward I's service from Savoy, the Alpine territory straddling what is now south-eastern France, western Switzerland and north-western Italy. His name is generally taken from the castle of Saint-Georges-d'Espéranche in the Viennois, where he worked on building projects for Count Philip of Savoy in the early 1270s.

That is roughly the extent of the biography. We do not have his date or place of birth. We do not know what he looked like, whether he could read, what language he preferred, or what he thought of Wales. He is a man visible almost entirely through accounts ledgers, and it is worth being blunt about that up front, because a great deal of what gets written about him is inference stacked on a very small evidential base.

How Edward found him

The Savoyard connection to the English crown was long-standing. Eleanor of Provence, Edward's mother, brought Savoyard relatives and retainers into English royal circles from the 1230s; Peter of Savoy, her uncle, held substantial English estates and gave his name to the Savoy in London. That network gave the English court decades of familiarity with Savoyard builders and their work.

The most commonly cited moment is Edward's journey home from crusade in 1273, when he passed through Savoy and stayed at Saint-Georges-d'Espéranche, then under construction. It is a tidy story and the timing works. Whether Edward personally met and recruited the mason there, or whether the arrangement was made through his Savoyard kin and officials, the record does not tell us. What is documented is that by 1278 Master James was in England, and from 1277 onwards a distinctly Savoyard set of building techniques starts appearing in the king's Welsh works — helicoidal (spiral) scaffolding-putlog patterns, particular window and latrine designs, and specific mason's marks.

What the role actually involved

James was not the architect of Caernarfon alone. He was, in effect, the director of Edward I's entire castle-building programme in Wales — titled in the accounts as magister operacionum regis in Wallia, master of the king's works in Wales. That meant setting designs, moving between simultaneous sites, hiring and directing masons, quarrymen, carpenters and diggers numbering in the thousands at peak, and answering to the exchequer for expenditure on a scale that strained royal finances.

It was a management job as much as a design one. The Welsh castles were built at speed, in hostile territory, largely by labour conscripted from English counties and shipped or marched in. Caernarfon alone consumed money at a rate that would today be measured in the hundreds of millions.

The other castles

His name is attached, with varying degrees of documentary firmness, to most of the great works of the second and third Welsh campaigns:

• Conwy (from 1283) — the tight, eight-towered rectangle on its rock, built alongside a walled town.
• Harlech (from 1283) — the concentric masterpiece on its crag, and the castle he later governed.
• Caernarfon (from 1283) — the most ambitious and the most deliberately symbolic of the group.
• Beaumaris (from 1295) — the last and most geometrically perfect, never finished.
• Earlier and lesser work at sites including Flint, Rhuddlan, Builth and Aberystwyth during the first campaign of 1277, plus later service in Scotland at Linlithgow.

A caution: "worked on" covers a wide range. For Harlech and Beaumaris the documentary link to James is strong. For Caernarfon it is good but less direct — he was master of the king's works in Wales while Caernarfon was rising, and the Savoyard technical fingerprints are present, but no surviving document says in so many words "James designed the King's Gate". Attributing every distinctive feature of Caernarfon to his personal hand goes beyond what the evidence supports.

Pay and status

This is where the record is genuinely rich, because the exchequer kept accounts and accounts survive. Master James was paid three shillings a day — a wage far above any ordinary craftsman — and from 1284 he received a lifetime pension of one shilling and sixpence a day, later raised, with provision for his wife Ambrosia to continue receiving payment after his death. He was granted land and a house at Caernarfon.

Set against the period, that put his annual income comfortably above that of many knights and in the range of the lesser nobility. The frequently repeated line that he was "paid more than most earls" overstates it — the great earls were vastly wealthier — but he was unquestionably one of the best-remunerated non-noble servants of the crown, and the pension arrangement for his widow signals real royal regard rather than routine contracting.

How much do we actually know?

Less than the confident tone of most accounts suggests. Almost everything rests on royal financial records, a handful of writs and grants, and the physical evidence of the buildings themselves. From those we can reconstruct where he was, roughly when, and what he was paid. We cannot reconstruct his intentions.

Even the Savoyard identification, now broadly accepted, was substantially established by twentieth-century scholarship — principally the work of Arnold Taylor — comparing building techniques and mason's marks across Savoy and north Wales. It is a strong argument built on careful comparison, not a signed statement of origin. A minority of historians have questioned parts of it. Anyone telling you exactly what James of St George believed about Roman imperial imagery, or precisely which stones he chose himself, is filling gaps the sources leave open.

Later life

In 1290 he was appointed constable of Harlech Castle — an unusual step for a craftsman, placing the man who built the fortress in charge of holding it. He served in that post for several years, and in the mid-1290s was drawn back into building work in the aftermath of the Madog ap Llywelyn revolt, including the start of Beaumaris. He also went north with Edward's Scottish campaigns.

He died around 1309; his wife Ambrosia was recorded as a widow by 1310. Where he is buried is not known. For a man whose work still dominates the skyline of half a dozen Welsh towns, that absence is a fair summary of how the medieval record treats even its most consequential builders.