Romanian is a romance language like Catalan, Italian, French, Portuguese, and Spanish so much of its core vocabulary is derived from Latin.

Why then even in core vocabulary does Romanian so often seem to be based on different Latin roots than its sister languages?

to go a merge < mergere

  • Catalan: anar < ambulāre
  • French: aller < ambulāre
  • Italian: andare < ambulāre
  • Portuguese: ir < īre
  • Spanish: ir < īre

Or is this perhaps a false impression I get just because these words stand out more? I'll add some more examples if I remember some. (I'm not immersed in Romanian anymore)

  • 7
    Oh, that's funny. Note that the French future stem of aller is ir- ("j'irai"). Many of the most common verbs are suppletive; I expect the verbs meaning to go in the other Romance languages to display some suppletion too.
    – Cerberus
    Oct 30, 2011 at 4:22
  • 2
    They do. en.wikipedia.org/wiki/Suppletion#Examples
    – dainichi
    Jan 29, 2012 at 5:40
  • 2
    Romanian infinitive of "to be" is fi, contrary to all other Romance languages where such word is based on esse, not fio.
    – iBug
    Dec 17, 2018 at 10:52
  • French marcher "to walk" is in line. There's probably some relation to marche "market", Etruscan merx, at least as adoption if not cognation; cp. travaille "work" and En. travel. marcher belongs to Ger. merken "remember, notice", En. mark; cp. Ger. Wegmarke "milestone, way point". Also cp. to score "to attain, acquire, mark". Also compare mercredi "wendsday" (went-day? lol), from Mercur "that god of messages and travel"
    – vectory
    Dec 25, 2018 at 20:58
  • 1
    French "aller" is from Latin. Especially when you conjugate this verb, you realize that it's a mix between 3 Latin verbs. Je vais, j'allais, j'irai = all the same "aller" verb. irai from ire. Vais from vadere, aller from ambulare.
    – Quidam
    Nov 20, 2019 at 2:04

7 Answers 7


Romanian is a language that's been well established in the Balkan Sprachbund (which is a German term linguists use to talk about a collection of languages that have shared features over a certain period of time).

Alongside Romanian in this Sprachbund are languages like Greek, Albanian and Macedonian (among others) and they have each played some influence in altering the other, and this is in all sorts of linguistic domains such as Phonology / Morphology / Syntax.

An interesting thing to note about Romanian is that, although deriving from Latin which had no articles, as East/West-Latin were still splitting up, Proto-Romance had started the grammaticalisation process resulting in the determiners observable today in Modern Romance. As Italian and Romanian both share the East-Latin branch they both originally had determiners that PRE-modified the elements they were attached to and over the course of time, due to influences from languages like Bulgarian (another Slavic language) and Albanian, these have their determiners AFTER the noun.

So to quote the oft-cited example here, look at the Italian word for 'the man', which is 'l'uomo' and then compare it with the Romanian word 'om' (man; notice the similarity) then after developing in this Sprachbund around other languages that put their word for the word 'the' AFTER the noun, Romanian also took this on and has 'the man' as 'om-ul'

So, as much as you do cross-linguistic comparison among the Romance languages, you will find out time and time again that Romanian is often the black sheep of the family due to its development in this language area where linguistic features are shared and different characteristics can be shown to have spread around to neighbouring languages after a long period of bi- and multi-lingualism throughout the history.

To swing this back to your point, this is exactly the reason why not just the Phonology, but also the word-stock (vocabulary) of Romanian is often very different from its sister languages in Western Europe. The same reasons for the sharing of features at a level of linguistic structure only happen in a very sort of 'linguistically intimate' setting and when two or more languages are close, the first thing to cross over are words into the other languages. So, if you see that structure has been shared, you'd be wise to make an educated guess that the level of vocabulary-absorption is considerably higher, and in this case you'd be right.

Romanian's history after splitting off and developing in a co-evolutative zone meant that it has slowly been moving further and further away from its core features in the Italic family, in not just words (but very much so), but also in many other aspects of structure (as I mentioned before).

I hope this helps you with your query.

  • 3
    A little correction: In italian "the man" (singular, you wrote "the men") is l'uomo, not *il uomo. :)
    – Alenanno
    Feb 6, 2012 at 9:36
  • Thanks for that! I've been speaking Italian for 6 years so I can't believe I didn't pick up on that myself!
    – Alxmrphi
    Feb 9, 2012 at 23:47
  • 1
    Incidentally, that's also the reason Romanian alone preserved cases, even though using them more in a fashion more like the Germanic and Slavic languages surrounding it than after Latin cases.
    – Joe Pineda
    Nov 27, 2013 at 13:06
  • 4
    To me this is mostly all correct but does not answer the question about choice of Latin vocabulary. Nov 28, 2015 at 4:16

Regarding Romanian's history: it's a member of the Eastern Romance (sensu strictu) subfamily or the "Vlach" languages. The peculiarities of Romanian are shared by its sisters in that branch such as Aromanian in Greece, Macedonia, Albania; Megleno-romanian in Greece, Macedonia, and Turkey; and Istrian in, well, Istria. Their peculiar history was influenced by at least 4 distinct factors:

  1. Being the earliest area to split off from the Empire, in 271 AD - leading to loss of stabilizing linguistic influence from Rome
  2. Geographical and later social separation from the other Romance languages, leading to less influence by and fewer borrowings from Germanic. E.g. 'white', instead of coming from Germanic blank, is alb from Latin albus
  3. The Balkan Sprachbund shared with Slavic, Greek, and Albanian. The Sprachbund had a significant influence on both grammar and, to a lesser extent, lexicon.
  4. The influence of the Paleo-balkan substratum which is not inconsiderable - some scholars report up to 500 roots may be inherited from Dacian

Of course, every modern Romance language has its own peculiar history, but the above I think covers the bases about what makes the Dacian language different.

  • The Balkan Sprachbund is present only in two Slavic languages (some may say one): Bulgarian and Macedonian. In Serbian those features are less present. So, that Bund cannot be Slavic. It cannot be Greek either, as the definite article after the noun is absent in Greek. That leaves the Bund centered on Romania, Albania, and the area between them, which coincides with the paleo-balkan substratum. But that Bund has not much to do with the question.
    – cipricus
    Nov 19, 2019 at 12:06
  • 2
    @cipricus I believe it is widely accepted that languages do not have to exhibit all the features of a sprachbund to be considered part of it, and Greek may be only a marginal member of the sprachbund for reasons like the one you mention, but just that one thing won't break the deal..
    – LjL
    Nov 20, 2019 at 1:41

Romanian does, in fact, have a reflex of Latin ambulāre: a umbla "to walk". (There are reflexes of ire as well, though it's become a future modal) I've often found that the familiar roots from other Romance languages do indeed exist in Romanian, though as in this case they'll have slightly different meanings.

The reasons for this are Romanian's relative isolation from other Romance languages, and its cultural domination by Slavic, Greek, and Hungarian neighbors for most of its history.


One possibility is because of its relative isolation from the other Latin-based languages. Imagine you're looking at early Latin dialects diverging from one another, and they've all got both ambulare and mergere.

If people in town A tend to use ambulare more, people in town B next-door might be influenced to do the same. But town Q isn't on the other side of the continent, and so there's no influence to end up with ambulare over mergere, so they've got 50/50 odds (assuming those are the only two words).

This theory requires two phenomena: 1) that a language variant's lexicon can drift over time and 2) that one language variant can influence a similar one nearby. I imagine neither of those would be controversial.

  • 4
    Yes that's the kind of explanation that seems natural but I was wondering if it's been researched etc since linguistics is so full of surprises that we don't naturally think of. Sep 27, 2011 at 20:31

An answer would be that Romanian is part of a different corpus of Latinized area of Europe than the rest of the surviving Romance languages. While the substratum of the first is Illirian-Thracian-Dacian (a vast linguistic area that we know little of, but that can be imagined as rather diverse but also constituted by related languages), that of the Western Romance is mostly Celtic, or related to the Celtic area and languages. That must have induced a different set of Latin roots being selected in the two areas. The Balkan Sprachbund (link in other answers) can be imagined as having a mainly neo-Latin period, before the coming of the Slavs.

While most of the Western part of Romance languages have survived the Germanic invasions, which were invasions by a minority of elite warriors who have been assimilated linguistically by the majority, the Eastern area suffered a different fate. The languages of the Germanic and Turkic peoples that invaded that part of the empire followed the same trend as in the West, as they were the same type of warriors. By contrast, the Slavs were very different. They came en masse, flooding by sheer number the Balkan region that previously spoke neo-Latin or proto-Romanian. They were able to assimilate even the Bulgars that came after them and to impose Bulgarian as a quintessential Slavic language.

Areas of Romance language survived though, especially in isolated mountainous regions (in the Carpathians and the Balkans), full assimilation in most cases taking place much later. To some extent Albanian can be considered a part of that area, as it has a huge Latin vocabulary, comparable and common to that of Romanian, although the Albanian verbal and pronominal structure was not Latinized. After the Second Bulgarian Empire, the Cuman-Vlach area north of the Danube became mostly Romance-speaking, while the Bulgarian-Vlach area south of the Danube became Slavic-speaking.

The relative uniformity of Romanian language north of the Danube (by contrast to the diversity of Iberian, Italian and French dialects) may indicate a sudden expansion of population in that area during the creation (and expansion to that area) of the Second Bulgarian Empire, as well as during the Hungarian stabilization of the area against the Mongols.


To be all fair and square to both sides of this discussion, the actual truth can go either way.

However, in defense of those who say that Dacian was already related to Latin, it makes more sense than you think, if you but try to think it over. This may explain the many dissimilarities between Romanian and other Romance languages, mainly why so many Romanian words have roots in Latin words different than their French, Spanish, Italian etc. counterparts. Roman Emperor Trajan himself said that he was from a Dacian (or related) village, where as a child he spoke a sort of Latin, much like all Dacians from other neighboring villages - this can be found in his correspondence and journals.

Of course that such a theory has a lot of opposition and I understand why that is so. If this theory held, it would refute 99% of all doctoral studies, books, and textbooks about Romanians' origins and their Roman-Dacian heritage. This would spell a big Ouch! for those who issued these studies. Campaigns to change the idea of Romanian heritage have taken place before, and not just in Romania. Before the 18-19th centuries, many sought to restore a Greater Dacia; after that, many sought to create a Greater Romania (which happened for a short spell of time). During Communism, they sought to teach Romanians of their Slavic heritage (thank the Soviets for that), although Romanian and Russian barely had anything in common, as opposed to the other Romance languages. After the 1989 fall of Communism, we returned to the idea of purely Roman heritage while seeking the support of the West and EU accession. I don't know what Dacians did wrong to be sidelined as an uncivilized people (as I, myself, was taught they were, in school). Interestingly enough, all invaders and foreign occupation have been described in Romanian history textbooks as evil, except the Romans. In terms of defining foreign occupation, what makes the Romans so much better than the Russians, the Poles, the Ottomans, or the Austro-Hungarians? The Romans were as much of an invading power as all the other ones. Do you see the politics in this? I know that I am biased, but I try not to be. In the end, all history is more or less biased, because we explain it in accordance with our current knowledge, perceptions, and feelings.

If we but apply common sense, here's one fact we are all aware of: The Roman Empire conquered only about 14% of the Dacian kingdom, which it occupied for only ~160 years. Comparatively, the Romans occupied other areas of Europe, Asia, and Africa for even 400-800 years at a time. For some reason, as stated by many historians, Dacians seem to have forgotten their language completely in the 160-year occupation of only 14% of Dacia, while other peoples who were under Roman occupation for many more centuries seem to have kept their language just fine. To me, that makes no sense.

The Dacian kingdom was pretty well developed when it was conquered (partially) by the Romans. Hence all the artisan work that has been discovered, the complex decorations on Dacian battle shields, Dacian fortifications and other ruins, the over 100 statues of Dacians exhibited in Rome (more than any other people conquered by the Romans), the 4 statues of Dacians guarding Constantine's arch, and the references to the Dacians made in the writings of Ptolemy, Socrates, Pythagoras, Herodotus and others as a well-organized, fearless people. They were well known at the beginning of our era as a strong kingdom that often plundered the Roman Empire before they were conquered, during Roman occupation, and even after the Aurelian retreat in 271 AD (although to a much lesser extent). After the Aurelian retreat south of the Danube they struck a deal with the Romans to protect the Roman Danubian borders from invaders (I'm not sure how well that played out).

References to Romanians' being referred to as Dacians date all the way to the 18th century. Horia, a famous Romanian revolutionary of that century, is depicted on a painting made for the Emperor of Austria at the time (whom he met at least three times at his court) as Rex Daciae (King of Dacia). Horia was known as Rex Daciae, being known to seek the reunification of old Dacia.

The idea that Romanians are pure descendants of the Romans was implanted in the minds of Romanians especially in Transylvania, by Transylvanian Romanian scholars who sought the support of the West, particularly France and the Vatican, in their struggle against Hungarian occupation and persecution of Transylvanian Romanians. It makes perfect sense to seek such a relationship with these European (and heavily Catholic) powers when the Catholic Church was clearly at odds with the Protestant movement (later on), the Eastern Orthodox Church, and the Ottoman Empire (the Muslim threat). The Orthodox Church was considered a Slavic church by most; the Slavs were not liked because of their connection to the great migrations that had heavily impaired European civilization and Catholic development, previously. Mind you, the pan-European movement to latinize language occurred mostly after the Great Schism (1054) between the Vatican and Constantinople. Therefore, the Dacians and proto-Romanians could not have been as heavily affected by Carolingian renaissance, or by the Catholic campaign of the Middle Ages that sought to instill the importance of Latin in the sociopolitical relations of Europe.

Moreover, the University of Hamburg, Germany, ran some paleogenetic tests - that is, they compared the DNA of bones dated up to ~6000 BC with the DNA of Romanians nowadays. The resemblance was staggering (around 75%). At the same time, they compared the DNA of Romanians with the DNA of Italians... only about 5-6% resemblance. There were more similarities between Romanians and Italians from Northern Italy, but nowhere near 75%.

Apparently, Micheál Ledwith, adviser to Pope John Paul II, having had access to the Vatican Library, stated that indeed Romanian is not resulted from Latin, but rather that both Latin and Romanian stem from the same mother language, Thracian. Basically, that means that Dacian and Latin were sister languages long before Dacians and Romans met. When Dacians surrendered to Rome, after years-long wars, there is no mention of any translators mediating the peace talks, which would have normally been the case. I conclude from this that Latin was a very important adstratum to the evolution of Romanian (as were the surrounding Slavic languages), but no more than that.

Either way, all of this can be interpreted both ways... There is a lot more to be told about these things, but I don't think I will. If you speak Romanian, you may want to watch these two documentaries released last year about the Dacians and Romanians' heritage. There are also versions on Youtube with English subtitles.

"Dacii - Adevaruri tulburatoare" - Documentary on YouTube 2012 HD

"Dacii - Noi dezvăluiri" - YouTube 2012 HD

PS - I am Romanian and I'm studying linguistics (particularly French). I am proud to be Romanian and am proud of my heritage. I don't dispute the Roman contribution to world civilization and to the Dacian people, by all means they were a great political and cultural empire. But I do dispute the idea that the main ancestors of Romanians are the Romans. For reasons aforementioned and others that I will not mention at this point, I do believe that there is more to Romanian heritage than the Romans and the Latin language. I believe that the Dacian and Latin languages were closely related, enough that there was never a complete latinization and subsequent replacement of the Dacian language by Latin. As for the Romanian words asserted to be Dacian, they are present in other Balkan countries/languages as well, and may very well be of Thracian descent (hence the pan-Balkan similitude). You will find specific sources to the things I just mentioned in the above documentaries (which you can find on Youtube with English subtitles).

  • 1
    Despite being almost a "wall of text" and having just a bit of admitted bias, I find this to be a good and interesting answer. There are many cases where the flow of language has taken a different path from the flow of genetics but the current lack of much knowledge of Dacian and Thracian language definitely leaves a gaping hole that is not easy fill in, especially with the many official theories pushed by dominating powers through history. Apr 8, 2013 at 2:36
  • 1
    Your post inspired me to look up some of the languages of the region and their history. It appears that there is really not enough information to come to any conclusion either way. My guess would be that the Roman influence started much earlier than the Roman occupation and that Dacians used Latin as a literary language rather than writing in Dacian. So maybe we can think of Dacian as a 'dialect' of Latin in the relaxed, mostly sociological sense that is used with 'dialects' of Chinese. Maybe the normal way to become a Dacian intellectual involved years spent in Roman settlements or even Rome.
    – user4938
    Nov 28, 2015 at 9:27
  • The relation between Dacian (if there was one single Dacian language) and Latin cannot be more than between two different Indo-European languages. This is the protocronist view on Dacians, the Romanian version of Protochronism. Latin was imposed by force even in Italy, were tens of very different languages were spoken.
    – cipricus
    Nov 19, 2019 at 12:15
  • That Dacia had one language (imposed in the same way as Latin was in Italy) before the Roman conquest is hard to imagine. The rapid Latinization of that area is more easily explained by the fact that Dacians were a federation of tribes with different languages (which explains why only exceptionally were politically united) and that linguistic unification came only with the Roman conquest, through Latin, in the same way as French and English have been adopted to this day by millions of Africans in less than 200 years.
    – cipricus
    Nov 19, 2019 at 12:17
  • For Romanian readers: take a look at this very funny but very interesting book: Dacopatia şi alte rătăciri româneşti
    – cipricus
    Nov 19, 2019 at 12:18

I would like to bring in a different point of view which was wrongly marginalized because it does not suit political agendas.

Romanian does not stem from Latin, this is why it is different compared to some of the Romance languages, it stems from Dacian. According to this view and ancient documents, the Dacian language (the language of the people that inhabited Romanian territory and the Balkans during Roman times) and the Latin language were related languages already, there are a lot of very strong scientific/historic arguments to support this idea.

An example: around 3000BC Aryans arrived in India and left there the Vedic Bible which is a monument to the Proto-European language that originated in the territory of Dacia. Well, in Vedic Sanskrit [water] is not [aqua] but is APA exactly like in Romanian APA that came directly from the Dacian language, and Sanskrit APA gave words that are similar to Romanian words that have no analog in Latin, ex. Vedic: APAYA, Rom.: APARAIA - a lot of water, flood.

  • 5
    This is complete fiction, presumably invented to suit political agendas. Romanian may well have some influence from the Dacian substratum, as Mark Beadles suggests, but in grammar and much of its lexicon it is solidly Romance. The derivation of apa from Latin aqua shows a sound-change also seen in patru, from quattuor and opt from octo.
    – Colin Fine
    Sep 4, 2012 at 0:10
  • 2
    It would help your case if you could include some references as well as the examples. For instance I can only find these other Sanskrit words for water and no "apa": जल (jal), अंब (aṃba), पानीय (pānīya). Sep 4, 2012 at 8:21
  • 1
    @hippietrail Well, āp(as) does appear in Vedic Sanskrit. For a related Indo-iranian example, consider "Punjab" < punch-ab, 'five waters'. But the alternation p-kʷ is common so I don't regard this as definitive. Sep 5, 2012 at 2:57
  • 1
    Persian آب (âb) / او (ow), from Middle Persian 𐭠𐭯 (āp), from Old Persian 𐎠𐎱𐎡𐎹𐎠 (ap-), from Proto-Indo-Iranian *hap-, from Proto-Indo-European *h₂ep-. The Sanskrit reflex is अप् (áp) and there is even a Latin reflex amnis (< *afni). (English amniotic doesn't seem to be related, Old English had reflexes but Modern English not apparently.) Romanian apă, on the other hand goes back to PIE akʷā- Sep 5, 2012 at 8:10
  • 4

Not the answer you're looking for? Browse other questions tagged or ask your own question.