Exactly

HOW WERE ROMAN ROADS BUILT?

Every book answers with four layers and four Latin names. The names come from a recipe for floors. What the Romans actually did was simpler, and smarter.

The Appian Way near the fourth mile, with wisteria in flower on a house beside the road. Photo: Kleuske, CC BY-SA 3.0, via Wikimedia Commons.

The short answer. Roman roads were built from the ground up, and from the water out. Surveyors laid out the line; workers dug a trench down to firm ground, filled it with compacted stone, gravel or earth, and finished it with a crowned surface that shed the rain into ditches on either side. Most roads were surfaced with gravel. The important ones, near cities, got fitted polygonal blocks of the hardest local stone. What made them last was not a secret recipe. It was a firm bed, good drainage and constant repair.

That is the honest short answer. The one you find in most books is tidier, and partly wrong.

The textbook answer

Open almost any book on Roman engineering and you will find the same cross-section: four layers with four Latin names. At the bottom the statumen, a bed of large stones. Above it the rudus, rubble bound with lime. Then the nucleus, a finer layer. On top the pavimentum, the paving. It is clear, it is easy to draw, and it is everywhere.

Two blocks of road cut open and drawn in perspective, in the manner of an old engraving. On the left, the textbook road between two kerbs: the polygonal paving stones are pulled back to show, from the top down, a pink layer of crushed tile, a layer of rubble in lime and a bed of large stones, numbered 1 to 4. On the right, a cambered gravel road in a shallow trench, its gravel pulled back to show one packed layer underneath, with a V-shaped ditch with water on each side and no kerbs, lettered A to C — Left, the textbook section, built from Vitruvius's instructions for floors. Right, the gravel road that excavations in northern Italy find most often. Schematic, not to scale.

Where the four layers come from

The names are real, and they are Roman. They come from Vitruvius, the only Roman architect whose treatise has survived. But they do not come from a chapter on roads.

In the seventh book of De architectura, Vitruvius explains how to lay a floor. First a bed of stones, none smaller than a fist. Then rubble mixed with lime, rammed down until it is at least three quarters of a foot thick. Then a layer of crushed tile and lime, six fingers deep. Finally the floor itself, slabs or mosaic, laid true and level. It is a recipe for a room.

In 1622 Nicolas Bergier, a lawyer from Reims, published a monumental history of the great roads of the Roman Empire, and read that passage as if it described a road. The idea stuck, partly because ancient texts on road-building are almost non-existent, and partly because for a long time nobody dug carefully through a Roman road to check.

Engraved portrait of a bearded man in a black gown with a stiff white collar, inside an oval frame lettered in Latin, set on a stone plinth that carries a five-line Latin verse — Nicolas Bergier, the lawyer from Reims who turned a floor into a road. Engraving by Jan van Vianen, 1699. Rijksmuseum, public domain.

What a Roman actually wrote

The only ancient text that describes the building of a road step by step is a poem. Around AD 95 Statius celebrated the Via Domitiana, the new road the emperor Domitian had built along the coast north of Naples, from the Appian Way at Sinuessa down to Pozzuoli. He lists the work in order. First, mark out the line with furrows and dig the earth out deep. Then fill the trench with other material and build up a bed for the crown of the road, ne nutent sola: so that the ground does not give way, and the stones do not sit on an uncertain bed. Then lock the road with kerbs on either side and wedges driven in close.

He also describes the crowd at work: felling trees, squaring beams, binding stones with lime and tufa, draining marshes, turning streams aside. And he gives the reason for all that effort. A journey that used to take a whole day now took barely two hours.

There are no four layers in Statius. There is a trench, a fill, a crown, kerbs and wedges. Not a recipe: a method.

Detail of a long, squashed coloured map: a crowned figure seated in a round medallion stands for Rome, and red road lines run out from it along a yellow strip of land between two bands of green sea, with small buildings and Latin place names along the roads — The Romans wrote little about building roads, but they mapped them. Rome on the Tabula Peutingeriana, a thirteenth-century copy of a late Roman road map now in Vienna: the roads leave the city as red lines, with the distances between stops. Konrad Miller's facsimile of 1887–88, public domain, via Wikimedia Commons.

What the excavations find

Archaeologists who have dug through Roman roads find something less tidy and more intelligent. In northern Italy, the archaeologist Michele Matteazzi brought the excavations together and found that roads outside towns were mostly gravel roads: one or two well-packed layers of gravel, sand or clay in a shallow trench, a gravel surface thirty or forty centimetres thick at the centre, a cambered profile, ditches on both sides and, often, no kerbs at all. Where the ground was wet, the builders improvised: packed amphorae at Parma, clay drains at Rimini, timber foundations near Ferrara. Elsewhere they raised the road on an embankment or cut it into the rock. In the north, paving stones are mostly an urban thing, and mostly from the age of Augustus onwards.

Sepia photograph of a long straight street between ruined walls and pillars, paved with polygonal stones, with high kerbs on both sides and three large stepping stones across it in the foreground — Paving was mostly a town thing. The Via dell'Abbondanza in Pompeii, with high kerbs and stepping stones for crossing, photographed by Giorgio Sommer between 1857 and 1875. Rijksmuseum, public domain.

The Romans themselves distinguished between earth roads, gravel roads and roads paved with stone, and the paved ones were the minority. Of the more than 400,000 kilometres of roads usually credited to the Empire, about 80,000 are reckoned to have been paved.

A street of large pale paving stones with grass in the joints runs towards a tall medieval stone house with an archway; on the left a church portico on columns above wide steps, on the right pink houses with green shutters and potted plants — Where the Appian Way crossed a town it became the main street. Here it runs through the old forum of Terracina, south of Rome, past the portico of the cathedral. Photo: Carole Raddato, 2014, CC BY-SA 2.0, via Wikimedia Commons.

The road built on its own stone

The road I know best is the clearest example. The Appian Way, begun in 312 BC by the censor Appius Claudius, was not paved at first: it was a gravel road. The paving stones came later, stretch by stretch. And on the first stretch out of Rome the builders hardly had to make a foundation at all. The road runs along the back of an ancient lava flow from the Alban Hills, hard, raised and straight, and its paving stones were cut from that same lava.

A map of southern Italy between the Tyrrhenian and the Adriatic. A dark red line leaves Rome at the top left, runs down the coast through Terracina, Formia and Minturno, turns inland through Capua to Benevento, then crosses the peninsula through Venosa and Gravina to Taranto and Brindisi. From Benevento an ochre line runs north-east through Troia, Canosa and Bari and down the Adriatic coast to Brindisi. Each station carries its Latin name and its modern one. A 100-kilometre scale bar — The whole road: the Appian Way from Rome to Brindisi, and the shortcut Trajan opened from Benevento in AD 109. The stations are the ones the ancient itineraries list; the lines join them in order and are not the exact route. Coastline and places from OpenStreetMap.

A road of big dark grey paving stones seen from close up, irregular in shape and worn smooth into hollows and ridges, with earth and pine needles in the wide joints, a line of cypresses and a few distant walkers at the far end — Lava paving on the Appian Way near the fourth mile from Rome, worn into hollows and ridges, with gaps where stones have gone. Photo: Nicholas Hartmann, 2019, CC BY-SA 4.0, via Wikimedia Commons

In 1756 Giovanni Battista Piranesi drew a section of it: ground rammed with stakes, a bed of lime, pozzolana and stone chips, and polygonal stones cut to a point underneath, driven into the bed like teeth into a gum. Even Piranesi hedged. About what some of the stones were for, he wrote "perhaps".

Etching of a paved road seen at an angle from above: a pattern of large polygonal stones between two rows of squared kerbstones, a small figure walking with a stick, trees on the bank behind, and a line of Italian lettering underneath — Piranesi again, this time on the surface: the paving and the kerbs of the Appian Way just outside Albano, from his Antichità d'Albano e di Castel Gandolfo, published in 1764. Rijksmuseum, public domain.

Why are Roman roads so straight?

Because they were laid out by surveyors with sighting poles and a groma, a cross on a staff for setting out straight lines and right angles, often from one high point to the next. And because the shortest line is the quickest way to move an army and its supplies. But they were not straight at any cost. They bent around hills, marshes and rivers whenever the ground said so, which is what good engineers still do.

A tall wooden staff with an iron spike stands on a black base in a museum hall; an arm at the top carries a horizontal cross with four cords and weights hanging from its ends — A groma, reconstructed at the LWL-Römermuseum in Haltern, Germany: cords with weights hang from the ends of a cross on a staff, and the surveyor sights along them. Photo: Arnoldius, 2014, CC BY-SA 3.0, via Wikimedia Commons.

A map with north at the top. A red line leaves the walls of Rome at Porta San Sebastiano, top left, and runs south-east past Domine Quo Vadis, the catacombs of San Callisto and San Sebastiano, the Circus of Maxentius and the tomb of Caecilia Metella. From the tomb it runs in an almost perfectly straight line through the green of the regional park, past the Villa dei Quintili, Casal Rotondo and Torre Selce, across the ring road, to Frattocchie at the bottom right. Grey roads, railways and Ciampino airport fill the countryside around it. A 2-kilometre scale bar. — A map of the Appian Way out of Rome: from Porta San Sebastiano past the catacombs to the tomb of Caecilia Metella, then almost dead straight for twelve kilometres to Frattocchie, at the foot of the Alban Hills. Roads, walls, park boundary and places from OpenStreetMap.

A narrow road of small dark square setts runs dead straight into the distance between umbrella pines and the tall rough cores of ruined tombs, with grass verges, a dirt path on the right and a power line against a clear evening sky — The straight, near Casal Rotondo, about six miles out of Rome. The line is Roman; the small square setts on the surface are much later. Photo: Livioandronico2013, 2014, CC BY-SA 4.0, via Wikimedia Commons.

So why did they last?

Ask anyone who has supervised roadworks what makes a road last, and you will hear two words: bed and water. A bed that does not settle, and water that never stays on the surface or soaks into the foundation. The Romans had both: the deep trench and compacted fill that Statius describes, a crown that sheds the rain, ditches that carry it away.

And they had a third thing we tend to forget. Rome had officials, the curatores viarum, whose job was to let out contracts for repairs. The Roman roads we walk on today are not proof of perfect construction. They are what is left of centuries of maintenance, on ground hard enough to keep them. Where the maintenance stops, you can see what happens in the photo at the top: the stones are still there, but the joints have opened and the water has got in.

Close-up of road paving: on the left, rows of small square grey setts; on the right, big rounded paving stones with pitted surfaces, one of them crossed by a worn groove — Two ages of the same road on the Appian Way in Rome: Roman paving stones, rounded by wear, next to a patch of small square setts laid much later. Photo: Livioandronico2013, 2014, CC BY-SA 4.0, via Wikimedia Commons.

Where to see one

Near Rome, the Appian Way beyond the tomb of Caecilia Metella keeps long stretches of its polygonal paving. Go on foot or by bike, and slow down on the stones: thin tyres slip into the joints. Since 19 September 2026 the first stretch, from Porta San Sebastiano, has been a limited traffic zone at weekends, and bicycles are exempt.

Seen from above at dusk, a tall white marble column with a carved capital and, beside it, the stump of a second one on its base, inside a railed enclosure on a paved square in front of a red-plastered palazzo — The other end. In Brindisi, above the harbour, the column long taken to mark the end of the Appian Way, and the base of its twin, which fell in 1528. Archaeologists still argue about what the columns were for. Photo: Stefn79 ph, 2021, CC BY-SA 4.0, via Wikimedia Commons.

…, Exactly? answers one question about a building, properly. If there is one you have always wondered about, reply to any Slow Letter and ask.

Sources: Vitruvius, De architectura VII.1; Statius, Silvae IV.3; Michele Matteazzi, "Ne nutent sola. Strade e tecniche costruttive in Cisalpina", Agri Centuriati 9 (2012); Nicolas Bergier, Histoire des grands chemins de l'Empire romain (1622); Giovanni Battista Piranesi, Le Antichità Romane, vol. III (1756). Translations are my own.