0—1mROMAN—2mrock
Turf & modern soil
An excavation in section · Hadrian's Wall

The Makers

Post-Roman / plough soil

Stone, mud, orders — and the layer where the work still lies.

The Roman horizon · AD 122

The band that remembers them — quarry chips, lime, hobnails, a centurial stone in the Wall's fabric.

Centurial stone
Quarry mark
Hobnail
Dressed sandstone · the curtain
Whin sill / bedrock
Evidence · Pattern · Inference · Reconstruction — faces of the makers: unknown
Roads · Wall · Tools · Hands · AD 122

Hadrian's Wall is authority becoming labour — an imperial order turned into surveyed lines, quarry faces, carts, lime, stone and weather by thousands of people whose names mostly vanished into the work. We can recover the work in extraordinary detail. We lose almost every worker.

How to read this page — the evidence behind every claim

E1 Direct evidence
E2 Physical pattern
S Specialist synthesis
I Engineering inference
R Reconstruction
Q Open question

The rule: the lower Wall is archaeology; the upper Wall and the human day are reconstruction.

Laser through the Wall
turf & stripped ground foundation E2 footing two faces E1 + rubble/clay core possible wall-walk possible parapet upper wall — not preserved (reconstruction) base ≈ 10 Roman ft / ~3 m (Broad)

If you could cut a clean slice through the Wall today, the lower fabric is archaeology; the upper finish is honest reconstruction.

From the ground up: a broad, shallow foundation of cobbles and clay; large footing stones; a stepped offset; then two dressed sandstone faces enclosing a rough core of rubble, clay and stonenot Roman concrete. E1 E2

Above that line — full height, any wall-walk, parapet or render — nothing survives. We show it ghosted, because that is exactly how certain we are. Q R

  • The recipe varied with the ground — bedrock on the crags, cobble rafts and turf on the Solway marsh.
  • Broad → Narrow: mid-build the planned ~10 Roman ft / ~3 m wall dropped to ~8 Roman ft / ~2.4 m — often sat on the wider foundation already laid. The reason is unresolved. S
  • Western sectors began as a Turf Wall, later rebuilt in stone.
The five-layer impact

Every stone placed had weight — on the land, and on the body that lifted it.

Layer 01

The land, opened

Before a stone moved, the army turned landscape into a measured military line: surveyors with cords and stakes, the turf and topsoil stripped to firm ground, a foundation cut to suit whatever lay beneath. The first act of building was disturbance — a scar drawn across farms and fells. E2

Layer 02

Quarry to haul

The Wall was made from the working landscape beside it — local sandstone wedged from faces often within a mile of the work, hauled short distances by ox-cart over bad ground. "I did not care to injure the animals while the roads are bad," wrote one frontier supplier. Long-haul stone was avoided by design. S

Layer 03

Build toil & precision

Two dressed faces, each stone (about a one- or two-man load) tapered back to grip a rubble-and-clay heart, raised in lifts behind timber working platforms. A century of roughly eighty men took a measured stint, sometimes marking it with a building stone — then leap-frogged ahead to the next. E1 I

Layer 04

The human cost

A cold, wet, windy line; heavy lifting; chronic illness on the frontier (one fort's report counts the sick, the wounded, the men with inflamed eyes). The day can be pictured, not timed — first light, orders, daylight work, beer and bread — but no AD 122 builder's diary survives. I R

Layer 05

Maintenance & legacy

The Wall was never finished once. It was repaired, narrowed, rebuilt, robbed and reused — kept alive by later hands across three centuries and quarried for farms and churches long after Rome. E1 S

The roster — names & lost names

A handful signed the stone. Thousands did not.

A name in stone is not a biography. These are the real traces — set against the silence they sit in.

Legio II Augusta
building legion
RIB 1638 · E1
Legio VI Victrix
building legion
RIB 1427 · E1
Legio XX V.V.
building legion
RIB 1852 · E1
Classis Britannica
fleet detachment — attested at Benwell
RIB 1340 · E1
name lost
Cocceius Regulus
centurion — a work-gang's stint
RIB 1862 · E1
name lost
Hrindenus
optio — quarry inscription
RIB 3331 · E1
Flavius Carantinus
a quarry mark — the closest we get to a hand
RIB 1442 · E1
name lost
Octavius & Candidus
supply — cash, hides, grain, bad roads
Tab. Vindol. 343 · E1
name lost

Names appear here as traces, never as portraits. RIB and Vindolanda references are used as evidence markers, not biographies.

A day before it was stone

The day can be pictured, not timed.

The morning probably began in the cold, by army habit: first light, a muster, orders given, the gangs counted and assigned. Builders likely used temporary camps, existing Stanegate bases and forward work sites — leather tents and military mess groups are a Roman-army reconstruction, not a surviving Wall diary.

What we have — diet, sickness and the supply world are real, from the Vindolanda tablets: bread, pork, beer, oysters, even pepper; and a strength report listing the sick, the wounded and the men with inflamed eyes.

Work likely ran with the daylight — long in summer, short and weather-broken in winter — with pauses shaped by weather, supply and command. The hardships are easy to believe: rain stalling earthworks, soft ground swallowing carts, strains and crush injuries from lifting stone all day.

What we don't — there is no AD 122 builder's clock: no recorded waking hour, break, shift or finish time. The diet and sickness evidence is from a Stanegate fort, mostly before the Wall — a powerful analogue, not a builders' record. Pictured, not timed.

And there was no culture of punished failure in the stones — what the archaeology shows is redesign, not collapse: the Wall narrowed, turf rebuilt in stone, forts added as the plan changed underneath the work.

Stone biography

What one stone could tell — if it could be traced.

Mostly, it can't. The Wall's sandstones are too alike to fingerprint a single block to a single pit. So this is a composite journey, not one real stone's life.

01

Quarried

Wedged from a local face — sometimes a quarry carries a Roman inscription, rarely a name.

E2
02

Dressed

Squared and chisel-drafted, cut to taper back and key into the core.

E2
03

Hauled

Carted a short distance over rough ground — most stone travelled in local sections.

I
04

Set

Laid into one of the two faces inside a century's measured stint.

E2
05

Repaired

Recut or refaced over centuries of upkeep — Gelt records quarrying for repair in AD 207.

E1
06

Reused

Robbed after Rome for farms, churches and field walls — the Wall's longest afterlife.

S

Composite stone biography. Most stones cannot be traced to one quarry — and none to one hand.

What the evidence can hold

What we have

  • The fabric: foundations, footings, offsets, two faces, the rubble-and-clay core.
  • The system: legions in sectors, centuries in stints, marked in stone.
  • The frontier's accounting mind — head-counts, sick-lists, supply letters.
  • A diet, a climate, a set of tools, a chain of command.
  • A few real names, and one quarry-hand: Flavius Carantinus.

What we don't

  • Whose hands shaped any curtain stone.
  • A real rota for the AD 122 build (our only rosters pre-date the Wall).
  • The human cost — no builder skeletons, no accident record.
  • The labourers' own voices, and whether local labour was levied.
  • The true height, wall-walk or finish of the curtain.
"Petra Flavi Carantini" — the rock of Flavius Carantinus.
A name cut into a quarry face by a labourer's own hand (RIB 1442). Not on the Wall, not a rank, not a face — but the closest we ever get to a maker.