The history of the Romanian language started in Roman provinces north of the Jireček Line in Classical antiquity, but there are 3 main hypotheses around its exact territory: the autochthony thesis (it developed in left-Danube Dacia only), the discontinuation thesis (it developed in right-Danube provinces only), and the "as-well-as" thesis that supports the language development on both sides of the Danube. Between the 6th and 8th centuries AD, following the accumulated tendencies inherited from the vernacular Latin and, to a much smaller degree, the influences from an unidentified substratum, and in the context of a lessened power of the Roman central authority, the language evolved into Common Romanian. This proto-language then came into close contact with the Slavic languages and subsequently divided into Aromanian, Megleno-Romanian, Istro-Romanian, and Romanian. Because of limited attestations between the 6th and 16th centuries, entire stages from its history are reconstructed by researchers, often with proposed relative chronologies and loose limits.
Background
A number of Romance languages were once spoken in Southeastern Europe for centuries, but the Dalmatian branch of this Eastern Romance disappeared centuries ago. Although the surviving Eastern group of Balkan Romance has in the meantime split into four major languages, their common features show that all of them originated from the same proto-language. Romanian, the largest among these languages, is spoken by more than 20 million people, primarily in Romania and Moldova. Aromanian has about 350,000 speakers who mainly live in the mountainous zones of Albania, Greece, and Macedonia. Some thousand people from the wider region of Thessaloniki speak the third language, which is known as Megleno-Romanian. The smallest Eastern Romance language, Istro-Romanian, is used by fewer than 1,500 speakers in Istria.
External history
Substratum
Little is known of the substratum language but it is generally assumed to be an Indo-European language related to Albanian. Most linguist like Kim Schulte and Grigore Brâncuș use the phrase "Thraco-Dacian" substratum, while Herbert J. Izzo and Vékony argue that the Eastern Romance languages developed on an Illyrian substrate. However, the small number of known Dacian, Illyrian or Thracian words excludes the systematic comparison of these idioms either with each other or with other languages. Dacian is represented by about a hundred plant names, 43 names of towns in Dacia as recorded by Ptolemy and around 1150 Dacian anthroponyms and 900 toponyms that have been preserved in ancient sources. The number of known Thracian or Illyrian wordsmainly glosses, place names and personal namesis even smaller.
Estimates of the number of Romanian words of substratum origin range between about 90 and 140. At least 70 of these words have Albanian cognates, which may indicate a common Albanian–Romanian substratum. However, borrowings from Albanian to Romanian cannot be excluded either. The linguists Gottfried Schramm, and István Schütz even propose that they were borrowed in several phases.
The largest semantic field (46 out the 89 considered certain to be of substratum) is formed by words describing nature: terrain, flora and fauna, and about 30% of these words with Albanian cognate are connected to pastoral life. The proportion of words with Albanian cognates are found in the semantic fields of the physical world (4.8%), kinship (3.2%), agriculture and vegetation (2.8%), and animals (2.7%).
Schütz argues that a number of Romanian words which are traditionally supposed to have been derived from hypothetical Vulgar Latin terms are in fact Albanian loanwords. Even Romanian words of Latin or Slavic origin seem to have been borrowed through Albanian mediation. Parallel changes in the meaning of a number of Latin words in the Albanian and the Romanian languages can also be illustrated. Recent studies however show that they are in fact substratum words in Romanian from Thraco-Dacian as evident in the rhotacism of intervocalic -l- in the pair Alb. vjedhullë - Rom. viezure for example, or the evolution of "dz" in words like bardzu typical of Latin to Romanian development. In general the argument that these are loanwords, based on the reason they are derivative forms in Albanian and only show as isolates in Romanian, is disproved by similarity of the Romanian word to Proto-Albanian, the language spoken before the 6th or 7th century. For example the Albanian word sorrë (crow) shows the change from ⟨t͡ʃ⟩ to s in Late Proto-Albanian, while Romanian has retained the old form cioară pronounced /ˈt͡ʃo̯a.rə/. A number of Albanian–Romanian calques exist.