Romania · Where the sources disagree

When and how Hungary took Transylvania

The map says 1000. The conquest actually took three centuries — and the only source describing who lived there first was written 300 years afterwards by a man inventing people to explain place names.

Updated 30 August 2026 6 min read One of 9 disputes in the Romania record

Two things are wrong with the date on this holding, and the second one is why people still shout about it.

The first is that "1000" makes a three-hundred-year process look like an event. Hungarian authority reached Transylvania in stages, moving west to east, and the stages are documented: the local ruler called the Gyula was removed around 1003, Ajtony in the Banat around 1028, the eastern mountains were not secured until somewhere between 1150 and 1200, and Maramureș in the far north was not organised as a county until 1303. Recording all of that as one holding starting in 1000 states the end of the process at its beginning.

The second is the question of who was living there when the Hungarians arrived. This is the other half of the Daco-Roman continuity argument, and it rests almost entirely on whether you can believe one book.

The book everything depends on

The Gesta Hungarorum — "The Deeds of the Hungarians" — was written by an anonymous notary of a King Béla, some time around 1200 to 1230. The Hungarian conquest it describes happened in the 890s. That is a gap of about three hundred years.

In it, the conquering Hungarians defeat a series of local rulers. One of them is Gelou, described as quidam Blacus — "a certain Vlach" — prince of the Vlachs in Transylvania. His land is rich in salt and gold-bearing rivers. His people are Vlachs and Slavs, armed with bows. They have been worn down by Cuman and Pecheneg raids. The Hungarian chief Tétény beats him at the river Almaș and he is killed near the Căpuș.

This is the earliest text that puts Romanians in Transylvania. Almost the entire argument follows from whether it can be trusted.

Why most Hungarian scholars say no

• The named enemies are wrong. Anonymus does not mention the opponents the Hungarians demonstrably did fight — Simeon I of Bulgaria, Svatopluk I of Moravia. Instead he names Gelou, Glad and Menumorut, who appear in no other source of any kind, anywhere.
• The book explains place names by inventing people. Scholars call it a toponymic romance. There is a village called Gilău; therefore there was a ruler called Gelou. The method is visible throughout.
• Pál Engel called these rulers half a dozen imaginary figures named after geographical features. István Bóna and Gyula Kristó agreed: Anonymus needed somebody for his conquerors to defeat, so he supplied them.
• C. A. Macartney argued the descriptive language is lifted from crusade chronicles — the world Anonymus actually knew — rather than from anything contemporary with the conquest.
• The place names are Hungarian. Most of the toponyms in the territory Anonymus assigns to Gelou are Hungarian in origin, which is awkward for a landscape he describes as Vlach and Slav.

Why many Romanian scholars say partly yes

• Alexandru Madgearu holds that Gelou was a real person and his name authentic, while accepting that Anonymus mixed oral tradition with invention. Victor Spinei argues most of the Gesta's reports have some real support even where anachronisms crept in. Ioan-Aurel Pop treats Gelou as one of three early tenth-century Romanian dukes. Tudor Sălăgean agrees the village name Gilău comes from the ruler rather than the other way round.
• Anonymus gets some things right. The earliest Hungarian rulers and some tribal chieftains are correct, which means he had access to something.
• The anachronism charge cuts less deep than it looks. Spinei's point about "Cumans" is that a chronicler using the current name for the current steppe power is doing something completely ordinary, not fabricating.

Dennis Deletant's summary has worn the best: the cases for and against the existence of Gelou simply cannot be proven.

What is in the ground

Both sides have gone to the archaeology and both have come back satisfied, which is usually a sign the archaeology is not deciding anything.

Hungarian arrival is well evidenced. Conquest-period cemeteries are real and published — Kolozsvár on Zápólya Street with twelve graves, Biharvár with eight, and others at Marosgombás, Magyarlapád, Dáva and Maroskarna. The graves contain sabres, bows, quivers, folded horsehides and harness; the women's graves contain pots of a type from the Don region, new to the Carpathian basin. They are spread fairly evenly across Transylvania rather than clustered at one entry point. Around the year 1000 a set of earthwork centres appears — Dobokavár, Kolozsmonostor, Tordavár, Hunyadvár — and three burials at Torda are dated by coins of Stephen I.

All of that establishes when Hungarians were there. None of it establishes who else was.

The pre-conquest question turns on a handful of sites, and Dăbâca is the test case. Florin Curta pointed out two pairs of bell-shaped pendants there resembling ninth-century Moravian jewellery — which would put a fortified centre on the ground before 900. Madgearu replied that pendants of that type occur only in sites from the last third of the tenth century to the first half of the eleventh, and concluded that a ninth-century fortress at Dăbâca is not yet proven. Cluj-Mănăștur, Moigrad and Ortelec are ambiguous or later.

The Hungarian excavators, for their part, record Slav funerary urns and graves of the Bulgar period underneath their own material — and say nothing at all about what became of the people who left them.

What would settle it

• A securely dated ninth-century fortified centre in Transylvania, published with its stratigraphy and its dating evidence before its ethnic attribution. That is the single find that would change the argument, and Dăbâca is where it has been looked for hardest.
• Any documentary reference to a polity in Transylvania between the fall of the Avar empire around 800 and the Hungarian arrival. Frankish, Byzantine and Bulgarian writers all had reason to notice one. None does. That is currently an argument from silence which a single charter would end.
• Ninth and tenth-century cemetery evidence with the burial rite described on its own terms. The trap here is severe: grave goods have been read as ethnic markers by both traditions for a century, and the readings disagree.

The part that is not in doubt

Strip out the contested question and a solid finding remains, and it is the one the map's single date obscures.

Hungarian authority did not arrive all at once. It took until about 1003 to remove the Gyula and 1028 to remove Ajtony. The east was organised only in the second half of the twelfth century, when the Székelys were settled as border guards — first in southern Transylvania, then moved east after the Saxons arrived. The Saxons got their charter, the Diploma Andreanum, in 1224; the Székelys had a royal official from about 1228. Counties are first attested from the 1170s, though the archaeology suggests a system existed in the eleventh century; the royal castle at Torda is documented in 1097 and Torda county in 1227. Maramureș became a county in 1303.

Three centuries. That is the finding. The year 1000 is the compression.

Sources

• Voivodeship of Transylvania — the administrative structure and its attestation dates.
• Gelou — the Gesta Hungarorum account and the debate, with the literature on both sides.
• Erdély története / History of Transylvania, ed. Béla Köpeczi (Hungarian Academy of Sciences), on the conquest-period cemeteries and the settlement sequence.
• Alexandru Madgearu, The Romanians in the Anonymous Gesta Hungarorum: Truth and Fiction.
• Gyula Kristó, István Bóna and Pál Engel for the sceptical position; C. A. Macartney on Anonymus's sources.
• Ioan-Aurel Pop and Victor Spinei for the case that the Gesta preserves genuine tradition; Dennis Deletant for the view that neither case can be proven.
• Florin Curta on the Dăbâca pendants.