Romania · Where the sources disagree
Rome held Dobruja for 550 years, and almost nobody counts them
Everyone knows about Roman Dacia, which lasted 165 years. Almost nobody mentions the stretch of Romanian coast Rome held for more than three times as long.
The short answer
Roman rule on the lower Danube ended when the frontier collapsed after 602, not with the Bulgar arrival in 680. The seventy-eight years in between belonged to nobody.
Here is a quiz question that catches almost everyone out. Which part of present-day Romania did the Roman Empire hold longest?
Not Transylvania. Rome was there for 165 years and left in the 200s. The answer is Dobruja — the flat, dry wedge of land between the lower Danube and the Black Sea, where the Greek ports were. Rome attached those ports to the province of Moesia in AD 46 and ran the place, continuously, until the frontier collapsed in the early 600s.
That is around 550 years. More than three times Roman Dacia. And it is missing from most popular accounts of "Roman Romania", which go straight from Trajan's Column to Aurelian's withdrawal and stop.
So why is this holding marked as contested? Not because anyone doubts Rome was there. Because the end date is routinely got wrong, and getting it wrong hides something real.
The two candidate endings
602 is the year the frontier army broke. The emperor Maurice ordered it to spend the winter camped north of the Danube instead of returning to quarters. The army refused, mutinied, marched on Constantinople and killed him. The Avar and Slav breakthrough that followed was never pushed back, and the Danube frontier was never rebuilt.
680 or 681 is the year Asparuh's Bulgars crossed the river, beat the emperor Constantine IV, and settled. A treaty followed. This is the date you will usually see.
They are seventy-eight years apart.
Why 602 is the better answer
Because the province stops functioning, and you can watch it happen:
- The forts go out of use across a band running from about 576 to 626 — a disintegration rather than a collapse on a single day.
- The frontier as an organised line is generally reckoned to have failed around 614 or 615.
- The coins stop arriving. The latest Byzantine issues found in Dobruja are two folles and two half-folles struck at Constantinople in 629 or 630, found at Mangalia — ancient Callatis. Imperial coin turning up in a province is the most reliable sign there is that an imperial administration is still paying for things. When it stops, the administration has stopped.
Put together: authority fails in a process starting in 602, is effectively over by the mid-610s, and leaves a residue on the coast for perhaps fifteen years more.
Why 680 keeps getting used
It is a better story. It has a named enemy, a battle, an emperor and a treaty, and it is the beginning of something rather than the messy end of something. Constantinople also never formally renounced the province — the emperors kept the titles, and the coast was reoccupied from time to time. On that reading the seventh century is an interruption rather than an ending, and 680 is when the interruption became permanent enough to require a treaty.
What is really being argued about
It comes down to what "held" means, which is the distinction this whole map is built on.
If holding a place means running it — officials, taxes, garrisons, money in circulation — then Roman Dobruja ends between 602 and the mid-610s, and the question is close to settled.
If it means claiming it in a way the neighbours eventually have to negotiate with, then it ends in 680.
The seventy-eight years between the two answers are the years when nobody was running Dobruja. That is a real historical condition, not a gap in the record. Dating the end at 680 quietly moves those decades into the Roman column and makes the imperial presence look continuous right up to the moment a named enemy takes over. It is tidier than what actually happened, and tidiness is doing the work.
What would sharpen it
More of this is answerable than in most of the disputes on this site, and the tools are coins and stratigraphy rather than texts:
- A published site-by-site coin sequence for Dobruja through the 600s, with the finds kept separate by context instead of pooled. The Mangalia coins of 629/630 are the current end point; one securely stratified later coin would move it.
- Excavation away from the coast. Nearly everything known about the seventh century here comes from Histria, Tomis, Callatis and Tropaeum Traiani. A frontier province is not the same thing as its four most-dug towns.
- Lead seals. Byzantine officials sealed their documents with lead, and the seals survive where the documents do not. They are direct evidence of whether anybody was still administering anything.
None of that will change the shape of the answer, which is that this province dissolved rather than fell. It would tell us which decade.
Sources
- Maurice's Balkan campaigns — the frontier war, the order to winter across the Danube, and the mutiny of 602.
- Theophylact Simocatta, on the campaigns of the 590s and the events of 602.
- Studies of monetary circulation in the Byzantine province of Scythia in the sixth and seventh centuries, for the Heraclian coins of 629/630 at Mangalia.
- Mihail Zahariade's work on Scythia Minor, for the late Roman provincial structure.