Number of Walled Cities in the Roman Empire

[Eleatic Guest]

1. Do we know how many Roman cities were walled?

2. And do we know how many had a second circuit wall?

3. And do we know how many times the first wall was breached in a siege but the second was held against the attackers?

The last question obviously cannot be answered by archaeology but it would be interesting to know if the building of a second wall was militarily worthwhile.

[Robert Vermaat]

1. Define 'Roman city'. Many if not all Roman towns had a wall, at least an earthen circuit, and I know of some towns (such as Aguntum in Austria), which had a nce wall but only on one side.

2. Honestly only Constantinople comes to mind.

3. Well.. all walls are eventually breached, so why not define the time you're thinking about? Rome had one wall which was never breached - the city fell when a gate was opened (410, 455, 546, 549, 1084). So you could argue that the building of gates is not military worthwhile.

[Eleatic Guest]

1. I also meant to include towns. My definition is circular: A town or city it is when it is walled, and if it is a town or city, then it is walled. :-)

2. Perhaps this was more common in the MA. I have this image of Carcassonne and Krak des Chevaliers  in my head.

3. Well, in the long run we are all dead...but it would be interesting to know how much time the existence of a second wall bought the defenders. And if the delay they produced was enough in some cases to save the city.

[Sean Manning]

Stefan, I don't know if there is an equivalent of Mogens Herman Hansen's Polis project listing all cities in the Roman empire, but most of the handbooks to Greek and Roman siege warfare will list you all the sieges for which we have detailed descriptions. You will want to re-read the technical manuals like Philon of Byzantion and Aeneas Tacticus too.

Most fortifications are concentric (even if that is just a ditch and a palisade) because the multiple layers have multiple benefits, like giving the defenders somewhere to herd their cattle away from wolves and robbers, protecting the gardens just outside the city walls from floods, and allowing twice as many defenders to shoot from a given length of circuit.

[Nathan Ross]

Do we know how many Roman cities were walled?

As Robert says, the vast majority of Roman towns and cities appear to have had walls at some point; particularly after the 3rd century in the west, even relatively minor towns in northern Gaul had them, often combined with a shrinking of the urban area, although in many cases it's unclear if they were still civilian settlements or had been turned into largely military installations.

Often it's hard to determine when the walls were built - the mura selenite of Bologna appear to date from some vague period between the late 4th and 6th century, although there were presumably walls there when Alaric besieged the town in 410.

Of all major western cities, I think only Carthage remained unwalled - apparently on purpose - until c.AD420.

how many had a second circuit wall?

This does seem to have been a largely eastern thing. Several cities may have had a proteichisma - Michael Whitby suggest that most major fortifications included one - but some of the larger cities in the east certainly did. Antioch had a double wall, and Nisibis had a double wall for at least part of the circuit and a stone moat as well. Resafa had a 'fore wall', and Procopius describes the defenders of Dara using the space between the walls to keep animals.

I can't think of any confirmed examples west of Constantinople though.One possibility might be Aquileia, although it's hard to say whether the different walls there were intended to replace each other, or to augment the circuit.

[Ildar]

This does seem to have been a largely eastern thing.

Yes, urban defenses were usually provided with a forewall (proteichismata) in Hellenistic times and in Late Roman times on the Eastern frontier of the Empire. The forewall was designed against highly organized and experienced in a siegecraft attackers.

Antioch had a double wall, and Nisibis had a double wall for at least part of the circuit and a stone moat as well. Resafa had a 'fore wall', and Procopius describes the defenders of Dara using the space between the walls to keep animals.

Also Amida, Singara, Resaina, Melitene, Martyropolis, Hatra, etc. For plans, cross-sections and elevations see Shelagh Gregory. Roman Military Architecture on the Eastern Frontier from AD 200-600. Amsterdam, 1997.

Defenses of the Tauric Chersonesus also had proteichismata. In the VI century AD, the Hellenicstic forewall was completely demolish, and then new forewall with towers was erected almost exactly in its place.

I can't think of any confirmed examples west of Constantinople though. One possibility might be Aquileia, although it's hard to say whether the different walls there were intended to replace each other, or to augment the circuit.

Do you mean Amm. Marc. XXI.12.13?

Then some with their shields raised over their heads, to be less hampered in fighting, others carrying ladders on their shoulders as before, rushed forward in fiery haste, exposing their breasts to wounds from many kinds of weapons. Still others tried to break the iron bars of the gates, but were assailed in their turn with fire or slain by great stones hurled from the walls. Some, who boldly tried to cross the moat, taken unawares by the sudden onslaughts of those that secretly rushed forth through the postern gates, either fell, if overbold, or withdrew wounded. For the return to the walls was safe and a rampart before the walls covered with turf protected from all danger those who lay in wait.

[Nathan Ross]

Do you mean Amm. Marc. XXI.12.13?

Thanks - I'd missed that reference! I was thinking only of the various plans of the city, showing at points in the circuit what appears to be concentric ramparts. But this passage of Ammianus does at least show that there was some sort of earthwork forewall in place in the mid 4th century.