Romania · Where the sources disagree
The Slavs north of the Danube, and when they stopped being Slavs
A fifth of the Romanian language is Slavic. One layer of those words can be dated to before the ninth century by a single sound change — and that is the sharpest evidence anyone has.
The short answer
The absorption was a long process, mostly between the 1000s and the 1200s. The end date of 1400 is an outer bound, not an event.
Somewhere between the seventh century and the thirteenth, a very large number of Slavic-speaking people lived in what is now Romania, and then stopped being Slavic-speaking. Nobody killed them and nobody expelled them. They were absorbed into a Romance-speaking population, and they left their vocabulary behind.
About a fifth of the Romanian language comes from Slavic. That is not a detail. It is the single largest outside contribution to the language, and any account of how Romanian came to exist has to explain it.
This holding runs from 680 to 1400, and 1400 is the number worth questioning.
What the date is, and is not
Nothing happened in 1400. There was no expulsion, no treaty, no last village.
1400 is an outer bound — the latest point at which you can reasonably argue there was still a distinct Slavic-speaking population north of the Danube. The best reading puts the bulk of the absorption between the eleventh and thirteenth centuries. The map carries the later date because a settlement does not end the moment the majority tips, and because choosing 1200 or 1250 instead would look more precise without being better supported.
There is a second thing this date is not, and it is the commonest confusion about the whole period.
Church Slavonic remained the written language of the church and the chancery in the Romanian lands until roughly 1600. That is two hundred years past the end of this holding, and it is not a contradiction. Church Slavonic was a learned written language maintained by clerks and clergy — like Latin in medieval England, which nobody takes as evidence that English villagers spoke Latin. Treating the two as the same thing is why the holding stops in 1400 and not 1600.
The part that can actually be dated
Most arguments in this period are about interpretation. This one has a piece of hard evidence, and it is rather beautiful.
South Slavic went through a sound change in the middle of the 800s called liquid metathesis: a particular sequence of sounds got reordered. Every South Slavic language today shows the change.
About eighty Romanian loanwords preserve the older, un-reordered shape.
Those words must therefore have been borrowed before the middle of the ninth century. No argument about where anybody was standing can move them. It is a fixed point.
Everything else layers on top of it:
- The oldest layer — those eighty-odd words, borrowed before the mid-800s.
- A large second layer from South Slavic and from Church Slavonic, after the change.
- A late, regional layer from individual languages, distributed exactly where later political geography would predict: Ukrainian and Polish words in Moldavia, Serbian in the Banat, Bulgarian in Wallachia.
And it is not only words. Romanian keeps a vocative case — the form you use when addressing someone directly — alone among the major Romance languages. Consonant clusters that early Romanian did not permit arrived with the borrowings.
The place names point later
Most of the older place names beyond the Carpathians are Slavic, and the sound shape of those names is datable too. They show forms borrowed after the eleventh or twelfth century; some match Middle Bulgarian forms from between the late 1000s and the early 1200s.
That is the strongest argument for a long, late absorption. Names get borrowed from neighbours, in whatever form the neighbour is using at the time. Slavic names entering Romanian in a twelfth-century shape mean Slavic speakers were still around, still being listened to, in the twelfth century.
The argument against reading it that way
Toponymy and living speech are not the same thing. A name can pass through one bilingual generation and then survive indefinitely in a population that no longer speaks the source language — the form is frozen at the moment of borrowing and says nothing about how long the donors lasted afterwards. On that reading, the twelfth-century phonology dates the borrowing, not the end of Slavic speech.
There is also no reason to expect a single answer for the whole country. Constantin Daicoviciu, in the 1969 Istoria României, split it: Transylvania Romanised early, Wallachia and Moldavia much later and under far heavier Slavic influence. If that is right, "when did assimilation finish" has different answers in different valleys and no single date is correct.
Where this turns into a fight
The archaeology of these centuries has been contested rather than discussed, and it is worth seeing why.
There is a material culture of the 700s to 1000s on both banks of the lower Danube, known in Romania as the Dridu culture after a site near Bucharest and recognisable by its burnished black pottery. Whose is it?
- Eugenia Zaharia (excavation report, 1967) and Ion Nestor (in Istoria poporului român, 1970) treated it as unambiguously Romanian. Nestor placed "proto-Romanians" as early as the sixth century and spread the formation of the Romanian people across the whole territory of the modern state.
- Ivan Bozhilov replied in 1970 that Dridu was Bulgarian, belonged to the First Bulgarian Empire, and had been dated to after 971 by Romanian scholarship specifically in order to detach it from Bulgaria.
- Constantin Daicoviciu, from inside Romanian scholarship, called it Slav or Slavo-Bulgar.
- Maria Comșa accepted a culture shared across both banks.
Lucian Boia's account of this episode is the one to read, because the argument was never really about pottery. Two national schools each needed the same eighth-to-eleventh-century material to be the ancestor of their own state, and the pots were being asked a question pots cannot answer.
The honest register belongs to Petre Diaconu, describing the population of eleventh-century Dobruja in the words of the Byzantine official Attaleiates who lived among them: mixobarbaroi, mixed barbarians — "un mélange valaco-bulgaro-petchénègue". A mixture is what the evidence supports.
What would settle it
- Dated cemetery populations rather than dated pottery — burial rite, grave goods and radiocarbon from the same graves, published before an ethnic label is attached. That would show when the two populations stop being distinguishable in the ground.
- Dating the Slavic place-name layer by sound-shape across the whole territory rather than region by region, which would test Daicoviciu's claim that absorption happened at different times in different places.
- Ancient DNA would date the mixing. It would not tell you which language anyone spoke, and this is an argument about language.
The likely outcome is not a date but a gradient — which is what the evidence already looks like.
Sources
- Slavic influence on Romanian — the proportions, the semantic fields and the three chronological layers.
- Old Church Slavonic — the liturgical and chancery language and its span in the Romanian lands.
- Eugenia Zaharia, Săpăturile de la Dridu (Bucharest, 1967); Ion Nestor in Istoria poporului român, ed. Andrei Oțetea (Bucharest, 1970).
- Ivan Bozhilov, "Kulturata Dridu i părvoto bălgarsko tsarstvo", Istoricheski pregled 26/4 (1970), 115–124.
- Petre Diaconu, Les Petchénègues au Bas-Danube (Bucharest, 1970), 104.
- Lucian Boia, History and Myth in Romanian Consciousness, 122–124.
- Roumen Daskalov, "Feud over the Middle Ages: Bulgarian-Romanian Historiographical Debates".