How Romanian packages meaning
Romanian grammar at a glance
At a glance
| Information type | Plain question | How Romanian encodes it |
|---|---|---|
| Actor | Who does the action? | SUFFIX |
| Recipient | Who or what receives the action? | SEPARATE WORDSUFFIX |
| Time | When does the action happen? | SUFFIXSEPARATE WORD |
| Aspect | Finished or still happening? | SUFFIXSEPARATE WORD |
| Certainty / mood | Real fact, or desired / possible / hypothetical? | SUFFIXSEPARATE WORD |
| Negation | Did it not happen? | SEPARATE WORD |
| Voice | Is the subject doing, being done to, or causing? | SEPARATE WORDSUFFIX |
| Evidentiality | How does the speaker know this? | SEPARATE WORD |
| Specificity | This particular one, or any one? | SUFFIXSEPARATE WORD |
| Quantity | How many? | SUFFIX |
| Possession | Whose is it? | SUFFIXSEPARATE WORD |
| Location / Space | Where? In which direction? | SEPARATE WORD |
| Information structure | What is this sentence about? What is new? | WORD ORDERSEPARATE WORD |
| Register | What social relationship does the speaker encode? | SEPARATE WORDSUFFIX |
How it stacks
One verb ending simultaneously encodes actor (who), time (when), and mood (real or hypothetical). Change the ending and all three change at once.
In synthetic tenses, the ending cannot be decomposed into separate morphemes for each feature — "vorbesc" is a single fused form. Compound tenses split the load: the auxiliary encodes person/number while the participle carries lexical meaning.
The enclitic definite article — a suffix on the noun — simultaneously encodes specificity (definite or not), number (singular or plural), gender, and case. This is unique among Romance languages.
The enclitic article fuses four properties: definiteness, case, number, and gender. Other Romance languages use a separate word before the noun (el, le, il). Romanian's postposed article is a Balkan areal feature shared with Bulgarian, Macedonian, and Albanian.
In detail
Actor in Romanian — Who does the action?
Packed into the verb ending as a suffix — the subject pronoun is optional and usually dropped. Four conjugation classes with different suffix patterns.
Subject pronouns (eu, tu, el/ea, noi, voi, ei/ele) are normally omitted — the verb ending alone identifies the actor. "Vorbesc românește" is complete; "Eu vorbesc românește" adds emphasis or contrast.
Romanian has four conjugation classes distinguished by infinitive ending: -a (a cânta), -ea (a vedea), -e (a merge), -i (a vorbi). Some verbs in classes I and IV take augments (-ez, -esc) in certain persons.
The 1st person singular and 3rd person plural forms are often identical (vorbesc = "I speak" or "they speak") — context or an explicit pronoun disambiguates.
Compare: English requires a separate pronoun — "I speak", "you speak", "they speak". French is intermediate — it writes pronouns but many verb endings sound identical, so the pronoun does the real work in speech.
Recipient in Romanian — Who or what receives the action?
A retained case system plus the preposition "pe" for animate/definite direct objects, reinforced by clitic pronouns that near-obligatorily double the full noun phrase.
"Pe" marks animate and definite direct objects — it is obligatory with proper names, pronouns, and kinship terms: "O iubesc pe Maria" (I love Maria). It is excluded with inanimates: "Văd casa" (I see the house).
Clitic doubling is nearly obligatory: the clitic pronoun (îl, o, îi, le…) appears alongside the full noun phrase it refers to. "Îl văd pe Ion" — both "îl" and "pe Ion" are present. This is a Balkan areal feature.
Clitic pronouns precede finite verbs (mă vede — he/she sees me) but follow imperatives (Spune-mi! — Tell me!) and can contract with auxiliaries (l-am văzut = îl + am văzut — I saw him).
Romanian retains a genitive-dative case distinction on nouns — the object form differs from the subject form for definite nouns: omul (NOM) vs. omului (GEN/DAT), casa (NOM) vs. casei (GEN/DAT).
Compare: Spanish has a similar "personal a" for animate objects but no surviving case system on nouns. Russian marks objects through case suffixes alone, with no need for a differential object marker like "pe".
Time in Romanian — When does the action happen?
Synthetic suffixes for present, imperfect, and pluperfect; a compound tense (auxiliary + participle) dominates for simple past; three competing future constructions range from literary to colloquial.
The perfectul compus (am vorbit — "I have spoken / I spoke") dominates everyday speech at ~90% of past tense usage. The perfectul simplu (vorbii) is largely regional (Oltenia) and literary.
Romanian has three future constructions: literary "voi vorbi" (voi + infinitive), colloquial "o să vorbesc" (o să + subjunctive), and colloquial "am să vorbesc" (am să + subjunctive). The literary form sounds formal or archaic in everyday speech.
The synthetic pluperfect (vorbisem — "I had spoken") is unusual — most Romance languages form the pluperfect analytically. Romanian retains this Latin-derived synthetic form alongside compound tenses.
Present tense covers both habitual and progressive meanings — there is no separate progressive form. "Citesc" means both "I read" (generally) and "I am reading" (right now).
Compare: French relies almost entirely on compound tenses for the past (j'ai parlé) and has lost the synthetic simple past in speech. Romanian is similar for the default past, but uniquely retains a fully synthetic pluperfect.
Aspect in Romanian — Finished or still happening?
Not independently grammaticalized — aspect is expressed through tense selection. The imperfect signals ongoing/habitual past; the perfect compus signals completed events.
Romanian does not have Slavic-style aspect pairs (perfective/imperfective verb roots). Instead, aspect is parasitic on tense: imperfect → imperfective reading, perfect compus → perfective reading.
In narrative, the imperfect provides background description (Ploua afară — "It was raining outside") while the perfect compus advances the plot (Am intrat — "I entered").
The presumptive mood is the one place where aspect shows up distinctly: "o fi citind" (might be reading — progressive) vs. "o fi citit" (might have read — perfective).
There is no grammaticalized progressive — "Citesc" covers both "I read" and "I am reading". Context, adverbs, or periphrastic constructions disambiguate when needed.
Compare: Russian has full grammaticalized aspect — nearly every verb exists as a perfective/imperfective pair (читать/прочитать = read.IMPF/read.PFV). Romanian, like other Romance languages, relies on tense distinctions instead.
Certainty / mood in Romanian — Real fact, or desired / possible / hypothetical?
Five moods: indicative (facts), subjunctive (să + verb, for desires/necessity), conditional-optative (aș/ai/ar + infinitive), presumptive (inference/hearsay — unique among Romance), and imperative.
The Romanian subjunctive uses the particle "să" + a conjugated verb form — NOT a separate set of verb endings like Spanish/French/Italian. Only 3rd person singular and plural differ from indicative: vorbește (IND) → să vorbească (SUBJ).
The subjunctive largely replaces the infinitive after most verbs: "Vreau să plec" (I want to leave), not "Vreau a pleca". This Balkan-type construction is shared with Greek, Bulgarian, and Albanian.
The presumptive mood (o fi bolnav — "he's probably sick") is unique to Romanian among Romance languages. It marks inference or hearsay and has present, progressive, and past forms.
The conditional uses invariable auxiliaries (aș, ai, ar, am, ați, ar) + infinitive: "Aș dori un ceai" (I would like a tea). Past conditional: "Aș fi vorbit" (I would have spoken).
Compare: Spanish and French have full separate subjunctive verb paradigms. Romanian's subjunctive is simpler morphologically (just a particle + mostly indicative forms) but used even more broadly — it replaces the infinitive in most contexts where other Romance languages still use it.
Negation in Romanian — Did it not happen?
The particle "nu" before the verb. Negative concord is obligatory — all negative words (nimic, nimeni, niciodată) co-occur with "nu" in a single-negation meaning.
Negative concord means "nu" is ALWAYS required, even when negative words like nimeni (nobody), nimic (nothing), or niciodată (never) are present: "Nu beau cafea niciodată" (I never drink coffee — literally "not drink coffee never").
Even when the negative word is fronted, "nu" stays: "Nimeni nu știe" (Nobody knows). This differs from Spanish, where fronted negatives drop "no" ("Nadie sabe").
In compound tenses, "nu" precedes the auxiliary: "Nu am vorbit" (I didn't speak), "Nu voi vorbi" (I will not speak).
Negative imperative: "Nu vorbi!" (Don't speak! — 2SG), "Nu vorbiți!" (Don't speak! — 2PL). The negative imperative form differs from the affirmative: "Vorbește!" (Speak!) vs. "Nu vorbi!" (Don't speak!).
Compare: French uses a split circumfix ne…pas around the verb. Spanish drops "no" when a negative word is fronted. Romanian is stricter — "nu" never drops regardless of position.
Voice in Romanian — Is the subject doing, being done to, or causing?
Active voice is default. Passive uses "a fi" + past participle. The reflexive "se" creates passive, reciprocal, and inchoative constructions. Romanian distinguishes accusative reflexive (se) from dative reflexive (își).
The analytic passive (a fi + participle) requires participle agreement in gender and number: "Cartea este citită" (F.SG), "Cărțile sunt citite" (F.PL), "Romanul este citit" (M.SG).
Romanian distinguishes accusative reflexive (se/mă/te) from dative reflexive (își/îmi/îți): "Se spală" (washes oneself — accusative) vs. "Își amintește" (remembers — dative, literally "reminds to-self").
The reflexive "se" also creates inchoative (change-of-state) constructions: "Vasul s-a spart" (The vase broke — no agent implied) vs. "Am spart vasul" (I broke the vase — with agent).
Romanian has no dedicated causative suffix. Causatives use "a face" + subjunctive or "a pune" + subjunctive: "L-am făcut să vorbească" (I made him speak).
Compare: Turkish stacks causative and passive as suffixes directly on the verb root. Romanian, like other Romance languages, uses separate constructions for each voice change.
Evidentiality in Romanian — How does the speaker know this?
Romanian has a grammaticalized presumptive mood — unique among Romance languages — that marks inference or hearsay. Lexical particles like "cică" (supposedly) add reportative force.
The presumptive mood is formed with the future auxiliary (o/oi) + fi + participle or gerund: "O fi citind" (might be reading — progressive), "O fi citit" (might have read — perfective).
Romanian occupies a middle ground: it has more grammatical evidentiality than French/Spanish/Italian (which have none), but less than Turkish or Quechua (which have fully obligatory evidential systems). The presumptive is grammaticalized but optional.
Additional lexical reportative markers include: "cică" (supposedly, they say — most common), "chipurile" (allegedly, often ironic), "pășamite" (apparently), and "se zice că" (it is said that).
The conditional mood can also carry a weak reportative flavor in journalism: "Ar fi plecat ieri" (He reportedly left yesterday).
Compare: Spanish and French have zero grammatical evidentiality. Turkish requires a choice between direct-witness (-di) and indirect/hearsay (-miş) in every past-tense statement. Romanian falls between — it has a grammatical tool (presumptive) but does not force its use.
Specificity in Romanian — This particular one, or any one?
An enclitic definite article attaches as a suffix to the noun — the only Romance language that does this. The indefinite article "un/o" is a separate word before the noun.
The enclitic article derives from Latin ille ("that"), the same source as Spanish el, French le, Italian il — but Romanian uniquely postposed it as a suffix: om → omul, casă → casa.
Article form varies by gender, number, and case: -ul (M.SG.NOM), -ului (M.SG.GEN), -a (F.SG.NOM), -ei (F.SG.GEN), -i (M.PL.NOM), -lor (PL.GEN). This fusion is captured in the noun ending fusion group above.
When a prenominal adjective is present, the definite article attaches to the LEFTMOST element of the noun phrase: "frumosul om" (the beautiful man — article on adjective) vs. "omul frumos" (the beautiful man — article on noun, adjective postnominal).
This postposed article is a Balkan Sprachbund feature shared with Bulgarian (-ът/-ата), Macedonian (-от/-та), and Albanian (-i/-a), despite these languages being from different families.
Compare: Spanish, French, and Italian all place the article BEFORE the noun as a separate word (el libro, le livre, il libro). Romanian is the only major Romance language to suffix it. Among European languages, only the Scandinavian languages (Swedish hus-et, Norwegian hus-et) share this postposed article pattern.
Quantity in Romanian — How many?
Noun suffixes mark plural, with complex patterns (-i, -e, -uri) that vary by gender and often trigger stem vowel changes and consonant palatalization.
Masculine plurals typically add -i with frequent consonant palatalization: băiat → băieți (d→ț before -i), copac → copaci. Feminine plurals use -e or -i: casă → case, mașină → mașini.
Neuter nouns behave as masculine in singular and feminine in plural — a unique three-gender system. "Trenul" (the train — M.SG article), "Trenurile" (the trains — F.PL article).
Stem vowel alternations are common: masă → mese (a→e), floare → flori (oa→o), carte → cărți (a→ă). These make plural formation less predictable than in Spanish or French.
Adjectives agree in gender and number: "om bun → oameni buni" (good man → good men), "fată bună → fete bune" (good girl → good girls).
Compare: Spanish plurals are simple (-s or -es with few exceptions). Romanian plurals involve three genders, multiple suffix patterns, and extensive stem changes — closer in complexity to German or Russian.
Possession in Romanian — Whose is it?
Possessive adjectives follow the definite noun (cartea mea — "my book"). A genitive article system (al/a/ai/ale) handles freestanding possessives. Genitive noun forms express noun-to-noun possession.
The possessed noun MUST be in definite form when using possessive adjectives: "cartea mea" (book-DEF my — correct), not "carte mea". Possessives agree with the possessed noun in gender and number: meu (M.SG), mea (F.SG), mei (M.PL), mele (F.PL).
The genitive article system (al/a/ai/ale) is unique to Romanian among Romance languages. It is required for freestanding possessives ("E al meu" — It is mine) and when the possessor follows an indefinite noun ("un prieten al mamei" — a friend of the mother).
For 3rd-person noun possession, Romanian uses genitive forms: "cartea băiatului" (the boy's book — boy-GEN.DEF), "cartea fetei" (the girl's book — girl-GEN.DEF). These forms merge with dative: genitive and dative are identical.
3rd person plural "lor" is invariable regardless of the possessed noun's gender: "cărțile lor" (their books), "casa lor" (their house).
Compare: Spanish uses possessive adjectives before the noun (mi libro) and "de" for noun possession (el libro de María). Romanian uniquely requires the definite article on the noun AND has the genitive article system (al/a/ai/ale) — a separate word class not found in other Romance languages.
Location / Space in Romanian — Where? In which direction?
Prepositions before the noun phrase: în (in/into), la (at/to), pe (on), din (from inside), de la (from a place). Many are compounds formed from simpler prepositions.
"La" serves double duty — static location ("Sunt la școală" = I am at school) and motion toward ("Merg la școală" = I go to school). Context and the verb's semantics disambiguate.
Romanian builds compound prepositions from simpler ones: din = de + în (from inside), de pe (from on a surface), de la (from a place), înspre (toward = în + spre). This is productive and transparent.
Some prepositions require genitive case on their complement: "deasupra orașului" (above the city — oraș-GEN.DEF), "în fața casei" (in front of the house — casă-GEN.DEF).
Romanian does not encode directional distinctions in the verb itself — the preposition alone carries the spatial meaning, unlike Japanese (iku/kuru) or Korean (-가다/-오다) where motion verbs encode direction relative to the speaker.
Compare: Finnish uses ~9 locative case suffixes on nouns (inessive, illative, elative, adessive, allative, ablative…). Romanian uses prepositions like other Romance languages, but uniquely combines them productively (din, de pe, de la) and governs case on some complements.
Information structure in Romanian — What is this sentence about? What is new?
Flexible word order marks topic and focus — topic elements move to the front. Clitic doubling is obligatory when objects are fronted, making it a grammaticalized topic-marking device.
Romanian word order is more flexible than French or Italian because verb agreement and case marking identify grammatical roles regardless of position. SVO, OVS, VSO, and SOV orders are all possible.
Clitic doubling is not just optional emphasis — it is grammatically obligatory when objects are topicalized or pe-marked. "Pe Ion îl văd" requires "îl"; dropping it is ungrammatical. This makes clitic doubling a true topic-marking device.
Subject-verb inversion can signal a question: "Scrie Elena o carte?" (Is Elena writing a book?) — or simply rising intonation on SVO order works too.
New/focused information tends to fall at the end of the sentence: "Am văzut-o pe Maria" (I saw Maria — Maria is the news). Old/given information clusters at the beginning.
Compare: Japanese has a dedicated topic particle は (wa) that explicitly marks the sentence topic. Romanian achieves similar results through word order + obligatory clitic doubling — the clitic resuming a fronted constituent is functionally analogous to a topic marker.
Register in Romanian — What social relationship does the speaker encode?
A multi-level T–V system with more distinctions than most European languages: tu (familiar), dumneata (mild polite), dumneavoastră (standard formal), and the reverential Domnia-Voastră.
Romanian has four levels of address — more than French's tu/vous or Spanish's tú/usted: tu (intimate/familiar), dumneata (mild polite — coworkers, acquaintances), dumneavoastră (standard formal — strangers, elders), Domnia-Voastră (reverential — very solemn contexts).
All polite forms derive from "lordship" (domnia): dumneata < domnia ta (thy lordship), dumneavoastră < domnia voastră (your lordship). This etymology parallels Spanish usted < vuestra merced.
Dumneavoastră takes 2nd person plural verb agreement even when addressing one person, but adjective agreement stays singular: "Dumneavoastră sunteți foarte amabil" (you-FORMAL be-2PL very kind-SG).
Among young Romanians, the shift to "tu" happens quickly. In professional contexts, "dumneavoastră" is the safe default until the other person invites informality. "Dumneata" occupies a middle ground that has no exact equivalent in French or Spanish.
Compare: French has a binary tu/vous distinction. Japanese has five or more levels with entirely different vocabulary and verb forms. Romanian's four-level system falls between these extremes — more nuanced than most European T–V systems.
Common questions about Romanian
Why does the Romanian definite article go on the end of the noun?
How does Romanian neuter gender work?
Is Romanian like Italian or Spanish?
Does Romanian have cases?
What's 'pe' doing before direct objects?
Sources for Romanian
The grammatical descriptions on this page are informed by the following published reference and descriptive grammars. Grammatical facts themselves are not subject to copyright; the scholars who documented them deserve attribution.
- Mallinson, Graham (1986). "Romanian." Croom Helm Descriptive Grammars.
- Pană Dindelegan, Gabriela (2013). "The Grammar of Romanian." Oxford University Press.
- Dobrovie-Sorin, Carmen & Giurgea, Ion (2006). "The Syntax of Romanian." Oxford University Press.
- Cojocaru, Dana (2003). "Romanian Grammar." SEELRC, Duke University.