Thracian language

• 1. Introduction

  ›

  Thracian reappears in the historical record in the year 570. A traveler named Antoninus of Piacenza visited a monastery in the Sinai. He recorded that its monks spoke five languages: Greek, Latin, Syriac, Egyptian, and Bessian, a dialect of Thracian. That detail is one of the clearest signs that Thracian outlived the Thracians as a distinct people. Thracian is classified as an extinct, poorly attested Indo-European language, once spoken across Southeast Europe by the Thracians. What survives of it today is scattered: a handful of inscriptions, a short list of glossed words, and place names spread across an ancient landscape. How much can actually be recovered from fragments this thin? And why do linguists still disagree about where Thracian belongs on the Indo-European family tree?

• 2. Words That Survived

  ›

  Ancient sources mention 23 words considered explicitly Thracian, each with a known meaning attached. None of the longer inscriptions found in Thrace has been read with confidence. Some may simply be jumbles of names or magical formulas rather than connected sentences. Enough Thracian lexical items have survived, though, to convince scholars that the language belongs to the Indo-European family. Thracian also survives indirectly, through personal names, toponyms, hydronyms, phytonyms, and divine names recorded by outsiders. A further, small set of words comes from Ancient Greek writers who explicitly labeled them as Thracian.

  Hesychius, a compiler of ancient Greek glossaries, preserved many foreign words, but only about three dozen of them are considered genuinely Thracian. Indo-European scholars have gone further. As they put it, even the notion that what the ancients called Thracian was a single entity is unproven.

  The word brytos, cited by many ancient sources, meant a beer made from barley. Rhomphaia referred to a broadsword, a term scholars have compared to the Latin rumpo, meaning to rupture. Skalme, cited by writers including Sophocles, Julius Pollux, and Marcus Aurelius, meant knife or sword. Zeira described a long robe worn by Arabs and Thracians alike, according to Herodotus and Xenophon. The word genton, meaning meat, appears in the works of Herodian, in the Suda, and in Hesychius. The longest of these engraved texts, cut into rings and stone across the region, have resisted every attempt at translation accepted by mainstream Indo-European scholarship.

• 3. Four Inscriptions

  ›

  A gold ring was unearthed in 1912 near the village of Ezerovo, in Bulgaria's Plovdiv Province. It carries the longest confirmed Thracian inscription, dated to the 5th century BC. The text runs across 8 lines in Greek script, with the eighth line wrapped around the rim of a rotating disk on the ring. The scholar Dimitar Dechev, whose name was Germanised as D. Detschew, separated the run-together letters into words and proposed a translation, though it has not won broad acceptance.

  In 1965, a second inscription surfaced near a village in Varbitsa Municipality, dating to the 6th century BC. Written in a variant of the Greek alphabet, it may be a tomb stele inscription similar to known Phrygian examples. The scholar Peter A. Dimitrov produced a transcription of it, but like the Ezerovo ring, it remains undeciphered.

  A third ring, found in Kaloyanovo Municipality beside the left hand of a skeleton, dates to the 5th century BC as well. It shows the image of a horseman, with an inscription wrapped around the figure, of which only 16 of an original 21 letters remain legible. One word, mezenai, is read as meaning Horseman. Scholars link it to the Illyrian term Menzanas, invoked in the phrase Juppiter Menzanas, meaning roughly Juppiter of the foals or Juppiter on a horse. Related words include Albanian mez for foal, Romanian minz for colt or foal, and Latin mannus for a small horse or pony. Gaulish manduos, meaning pony, is even thought to lie behind the tribal name Viromandui, meaning men who own ponies.

  A damaged stele turned up during excavation of the Sanctuary of the Great Gods on the island of Samothrace. Its inscription dates to the late 5th or 4th century BC. The final two lines may belong to a Greek postscript, but the rest of the text is clearly not Greek at all.

  Those four inscriptions are the only substantial Thracian texts ever recovered. But the language left many more traces in the names of towns and rivers across the region.

• 4. Traces on the Map

  ›

  Thracian, or a closely related set of dialects, was spoken across what is now Bulgaria, Romania, North Macedonia, and Northern Greece. Its reach extended into European Turkey and parts of Bithynia, in Northwestern Asiatic Turkey.

  Many town names across ancient Thrace and Dacia end in suffixes such as -dava, -daua, -deva, -deba, -daba, or -dova, all meaning city or town. Farther south, the common endings shift to -bria, meaning town or city, and to -disza, -diza, or -dizos, meaning fortress or walled settlement. Other southern names end in -para, -paron, -pera, or -phara, all referring to a town or village. The geographer Strabo translated -bria as polis, the Greek word for city-state, though modern scholars aren't fully convinced that rendering is accurate.

  The -disza, -diza, and -dizos endings trace back to a root meaning to knead clay. That same root produced meanings such as making bricks, building walls, or simply a wall. Linguists call this shift satemization, the same sound change that turned an old PIE *gh- into a hissing consonant in Thracian. Cognates include the Ancient Greek word teichos, meaning wall or fortified town, preserved in the town name Didymoteicho, and the Avestan word daeza, meaning wall.

  Some scholars trace the -dava ending specifically to the Dacian language, with the rest of the suffixes coming from Thracian proper. Yet towns with -dava endings turn up as far south as Sandanski and Plovdiv, well outside any strictly Dacian territory. Some -dava names, such as Pirodava, even share features with -diza names like Pirodiza, blurring the supposed divide.

  Roman sources contain the first written mentions of the name Dacians, and the geographer Strabo identified the Daci as the Getae, a Thracian tribe. Writers including Dio Cassius, Trogus Pompeius, Appian, Herodotus, and Pliny the Elder all treated the Dacians and Getae as Thracians who spoke the same language. That consensus is why Dacian is usually classified as a dialect of Thracian rather than a separate language. It's also possible -dava and -bria simply described two different kinds of settlement in the same language, urbanized towns versus more rural, steppe-like ones.

• 5. A Disputed Family Tree

  ›

  Most modern linguistic textbooks place Thracian in its own branch of Indo-European, or group it with Dacian as a joint Daco-Thracian branch. Older textbooks often grouped Thracian with Illyrian or with Phrygian instead, though the Phrygian connection has mostly been abandoned.

  The linguist Harvey Mayer proposed a different classification altogether. He argued that Thracian and Dacian belonged to the Baltic branch of Indo-European, or at least sat closer to Baltic than to any other branch. Most linguists still don't accept that idea, and it remains a minority position among specialists.

  A more fringe hypothesis holds that Thraco-Dacian formed its own branch of Indo-European alongside Baltic. That idea runs into a wall because the linguistic community overwhelmingly accepts a unified Balto-Slavic branch, a grouping this fringe theory would have to break apart. None of these classification debates address a separate, harder question: how and when Thracian actually died out as a spoken language.

• 6. The Language's Long Fade

  ›

  The 19th-century Greek educator Vlasios Skordelis argued that Alexander the Great's subjugation of the Thracians led them to fully assimilate into Greek culture. In his view, they became as Greek as the Spartans and Athenians, and he even considered the Thracian language itself a form of Greek.

  Writing in 1997, the historian Crampton concluded that most Thracians were eventually Hellenized or Romanized, with isolated pockets surviving in remote areas until the 5th century. Marinov, writing in 2015, pushed the timeline later, arguing Thracians were likely fully Romanized and Hellenized after the last references to them in the 6th century.

  Not every historian agrees: Trever, writing in 1939, held that the interior of Thrace was never fully Romanized or Hellenized. Slavonization followed after that.

  The historian Weithmann, writing in 1978, described a Thracian and Dacian population that incoming Slavs found only superficially Romanized. That population had never strongly identified with Imperial Rome, while the Greek and Roman population, mostly soldiers, officials, and merchants, either fled or were killed.

  The old name Pulpudeva survived that upheaval, living on in Slavic languages as Plovdiv rather than the Greek name Philippopolis. Some scholars point to that survival as evidence Thracian was not completely wiped out even by the 7th century.