The Turkish Influence on the Romanian Language


The Turkish language (called also Ottoman or Osmanli) has had, during almost five centuries, a considerable influence on Romanian, from the end of the 14th to the middle of the 19th century, and having some local, dialectal extensions in Dobrudja until nowadays. The number of the lexical loans amounts to over 2.750 words. Their inventory, of which almost two thirds disappeared from the today’s Romanian language, does not include those Turkic loanwords that are not Ottoman, but pre-Osmanli Turkic languages such as Pečeneg, Kouman or Tatar languages, and those Turkish words which entered into Romanian by way of other (mostly Balkan) languages, such as Greek, Bulgarian, Serb, Gipsy, or Hungarian. The Turkish linguistic influence on Romanian was the result of a historical context in which the Ottoman Empire extended its domination and suzerainty on the Romanian Principalities, and exercised a longstanding influence on their social, administrative and economic life until the second half of the 19th century.

1. Studies on the Turkish influence have a long tradition in Romania, starting with the first years of the 18th century (Dimitrie Cantemir), but the work of Lazăr Şăineanu, Influenţa orientală asupra limbei şi culturei române [The Oriental Influence on the Romanian Language and Culture], Bucharest, 1900, has remained for more than a century, the only monographic synthesis on the subject. Despite its undeniable merits, this work fatally became incomplete and obsolete, as a result of the lexicological and lexicographic progresses, of the publishing of new documentary sources (scientific editions of old texts, dialectal glossaries and atlases, Romanian and Turkish historical and dialectal dictionaries, etc.), as well as of the gradual change of the status of many words in modern Romanian. The history and the actual situation of the Turkish loanwords in the medieval and contemporary Romanian language are largely described in our recent work, Influenţa turcă asupra limbii române. I. Studiu monografic. II. Dicţionarul cuvintelor româneşti de origine turcă [The Turkish Influence on the Romanian Language. I. Monographic Research. II. Dictionary of the Romanian Words of Turkish Origin], Bucharest, 2009, 2010.

2. The Turkish influence can be distinguished by some particularities, as follows:

a. The source of the loanwords cannot be solely Turkish, but from some other neighbouring languages as well (multiple etymology), since many Turkish loanwords can be found in different South-Eastern or Eastern European languages. On the other hand, in many cases the source is not the modern Turkish standard language, but one of the archaic, popular or dialectal variants of Turkish: e.g. Rom. musafir ‘guest’ < Old Turkish müsafir (cf. Modern Turkish misafir).

b. The Turkish words penetrated into the Romanian language either as official, possibly cultural loanwords, thanks first of all to the individual bilingualism of the Romanian aristocratic and erudite class (although the majority having remained at the level of personal borrowing or jargon), or on the popular, oral way of direct contacts between common people. The proportion of these two categories is approximately equal, but the geographic dispersion and the viability of the popular borrowings have been much higher.

c. The discontinuity and disparity of the attestations or the different forms and meanings of some words are proofs of the fact that they were borrowed repeatedly (e.g. amanet ‘pawn’; 18th century: emanet;19th century: amanat, 1594).

d. The temporal stratification of the Turkish loanwords of the Romanian language can be studied from three points of view: chronological, viabil and synchronical status of the words in different phases of the influence:

– from the end of the 14th century to the 15th century, with 23 loanwords, all of them having survived until the 20th century;

– from 16th century: 152 loanwords, with their majority, 66%, preserved until today;

– from 17th century: 502 loanwords; 52% of them are found in today’s Romanian, but only 168 are still usual, the rest became obsolete.

– 18th century is marking the top point of the Turkish influence, with more than 1.000 loanwords, but only 274 of them are elements of the usual vocabulary of the today’s Romanian speakers;

– 19th century, with more than 1.000 words, most of them borrowed until the middle of the century, when the Turkish influence on Romanian language as a whole came to an end, and a strictly dialectal influence, mainly in Dobrudja, became dominant; 417 of the 19th century loanwords can be found today in the usual vocabulary, and 161 are in course of disappearing.

e. As far as the territorial repartition of the loanwords is concerned, the great majority are concentrated in the southern and eastern historical provinces of Romania: in Dobrudja (where the dialectal borrowing was dominant, due to the fact that this province effectively belonged to the Ottoman Empire and has had a consistent Turkish population), in the Banat region (which was for two centuries under Ottoman administration, but has never had a significant Turkish population), and in Moldavia, Wallachia and Oltenia (provinces which had for centuries a status of vassalage and used to consciously copy the administrative structures, the economic features and the mores of the Ottoman Porte). As a result, there are numerous loanwords, sometimes with different phonetic features, common either to Moldavia, Wallachia and Oltenia, or to Wallachia and Dobrudja, or either to Oltenia and the Banat, or to Dobrudja and Moldavia. Many other terms are specific only to one of these provinces, even to the Banat (where most of the words of Turkish origin were borrowed through the Serbian language). Transylvania, where the administration was Hungarian, is the only province whose Romanian vernacular generally escaped from the Turkish influence, even if there are a few local loanwords (aga, hăzap), and if, later on, a lot of literary and popular terms of Turkish origin expanded therein from the southern and eastern parts of Romania.

f. We can also remark a stylistic repartition of the words in the moment when they were borrowed, placing them in one of the functional  (colloquial, official, narrative, scientific, professional, familiar, vulgar, etc.) variants of the Romanian language, as well as several later transfers from one registry to another, sometimes due to a semantic evolution.

3. All the compartments of the Romanian language have traces of the Turkish influence. Nevertheless, there are no longstanding and consistent effects but in the vocabulary. The innovations in phonetics and grammar are usually linked only to some changes in the distribution of the elements, but not in their inventory.

a. In the mass of the more than 2.750 (according to our accounts, 2.760) Romanian terms borrowed from the Turkish language, predominant are the nouns (91%) and the adjectives (6.3%); there are also 32 adverbs, 37 interjections, six verbs, two conjunctions, one pronoun and one preposition. These loanwords have sometimes preserved the form of the etymons, but there are also some cases of phonological and/or morphological adaptation. The reflexes of the Turkish vowels ö, ü and of the very specific spirant ğ, which don’t exist in Romanian, are (i)o, (i)u or i,respectively g, h or zero (e.g. Romanian ghiol ‘lake’ < Turkish göl; Romanian dulgher ‘carpenter’ < Turkish dülger; Romanian ghiveci ‘flower pot’ < Turkish güveç; Romanian iama ~ iagma ‘havoc’ < Turkish yağma);the groups /čï/,/ǧï/,/šï/ and//, unpronounceable in the standard Romanian, changed their vocalic features to i or e, and at the end of the words, by morphologic adaptation, to iu or ie (Romanian arşic ‘knucklebone’ < Turkish aşık; Romanian saşiu ‘cross-eyed’ < Turkish şaşı); the Turkish long or double vowels and the geminated consonants have been generally reduced (Romanian adet ‘custom’ < Turkish âdet; Romoanian ursuz ‘morose’ < Turkish ūrsuz < uğursuz; Romanian ghiulea ‘cannon ball’ < Turkish gülle); the /t/, /d/, /č/, /ǧ/ and /š/ in front of an affricate consonant have changed by total or partial dissimilation (Romanian zarzavagiu ‘greengrocer’ < Turkish zarzavatçı). Many loanwords have undergone different alterations of their absolute end, with the only scope to be adjusted to the Romanian flexional types; the vocalic endings under accent, unusual in Romanian, have been adapted as follows: Turkish -ı/-i/-u/-ü > Rom. -iu or -ie (Romanian burghiu ‘auger’ < Turkish burgu; Romanian muşteriu ‘customer’ < Turk. müşteri; Romanian sarailie ‘almond cake’ < Turkish saraylı); Turkish -a > Romanian -a or -ă (Romanian balama ‘hinge’ < Turkish bağlama; Romanian ciorbă ‘sour soup’ < Turkish çorba); Turkish -e > Romanian -ea (Romanian cişmea ‘water pump’ < Turkish çeşme). Many other words have been modified phonetically as a result of reinterpretations induced by the general or dialectal rules of the Romanian phonological and morphological systems, by analogy or by phonetic accidents. These new forms sometimes alternate with the etymological, ‘correct’ ones. It can be stated that in the domains of phonetics and morphology there are no real Turkish borrowings, but only changes in the distribution of several phonemes (e.g. the consonant h)and phoneme groups, as well as of some nominal flexional types.

b. The syntax, even if more permeable, was also superficially influenced. There are nevertheless several relational elements (ama ‘but’, başca ‘besides’, ioc‘not at all’)and syntactic structures borrowed from the Turkish language. Some of these were transposed to Romanian as compounds (get-beget < Turkish cet becet)or, on the contrary, were simplified by lexical-semantic condensation (Romanian bidinea ‘whitewashing brush’ < Turkish badana [fırçası]), in this way quitting the syntactic field. Other structures are not borrowings, but imitations of Turkish models, reproduced by lexical or phrasal calque (e.g. Romanian ficat-alb ‘lung’ (cf. Turkish akciğer ‘id’, both, word for word, ‘white liver’). Some word-formations are the result of false syntactic interpretations (contraction, agglutination, deglutination).

c. The Turkish loanwords belong to a large sphere of semantic spheres, covering almost every field of human life and activity, and reflecting the extended influence of the Ottoman material and spiritual civilization on the Romanian one. The proportions of this influence were nevertheless not equal in all fields. The richest terminological groups of Turkish loanwords are those related to the society (social and administrative organization, armed forces, finances, justice, etc. (33% of the loanwords), to the human being (his domestic universe, physical and psychological status, manner of clothing, lodging and nourishing (24%) and to the human action onto the nature (the domains of production, of work (15%). The environmental areas rank medium (fauna, flora, other realities of the environment), the particularities of the surrounding realities and their perception by man (colors, dimensions, quantities, modalities, etc., expressed by adjectives, nouns, adverbs, interjections, etc.), but the abstract notions rank lower The poorest terminological groups are those linked to culture (amusements, music, folklore, education, literature, sciences) and to human relations as they rank lowest.

If one wants nevertheless to compare these terminology groups from the point of view of the loanwords having survived until today, it will be clear that, even if every group had lost a lot of words, the terms related to the society fell to the last position, with the heaviest losses. From the viewpoint of the viability of the terms, i.e. of their effective value, it is suggestive that today’s Romanian retained the following proportions from the total amount of the loanwords: nature (66%), human action onto the nature (2.3%) particularities of the surrounding realities (57.6%), human being (55%), abstract notions (48.5%), culture (47%), human relations (38.7%) and society (16.7%). The group of terms referring to social realities, which had gained the greatest number of Turkish words, was finally the least viable. In the same time, more than three fifths of the terms related to natural realities and to professions have survived until today. Needless to say that many words, after being borrowed, were subject not only of phonetic changes, but of different semantic evolutions as well (extensions, generalizations, depreciations, restrictions, analogies, etc.).

4. The cultural influences, the longstanding direct contacts between Romanian and Turkish people, their cohabitation in one or another area, as well as their ethnic mixture in certain historical periods and at certain social levels, have made possible that, beside the common terms, several place-names (Adacale, Bender, Bugeac, Crâm, Siutghiol, Tuzla a.s.o.), as well as names of persons(e.g. Aslan, Caraman, Deşliu, Ghelmez, Lehliu, Magearu, Mungiu), animals (Barac, Hormuz, Samson), stars (Ciobanul, Tereziile), etc. were borrowed from the Turkish language. Other names have been formed in Romanian on the basis of common terms of Turkish origin.

5. Word-formation was the linguistic field having the greatest profit following the lexical borrowing from Turkish. The Romanian language re-used the borrowed lexical material in two manners: on the one hand, several suffixes (-iu, -giu, -liu, -lâc)and one composition element (baş) were detached from the structure of the loanwords and became functional, that is, productive in Romanian, being attached to words of other origins than Turkish (e.g. argintiu < argint < Latin; bragagiu < bragă < Russian, duelgiu < duel < French, Latin ; savantlâc < savant < French; baş-răzeş< răzeş < Hungarian. On the other hand, new words and variants were created on the basis of the loanwords, either with the own formative means (by derivation, composition or conversion), or by reinterpreting the Turkish etymons or the borrowed forms through derivative analogy, popular etymology, contamination, etc. Without having borrowed a great number of Turkish words, it wouldn’t have been possible to form in Romanian, on their basis, more than 2.000 new words, as well as many locutions and expressions.

6. Our statistics done on a great amount of texts and lexicographical sources, and aimed at revealing the frequency and functional value of the lexical borrowings, lead us to the conclusion that in the present-day Romanian language there are about 1.250 words of Turkish origin, of which 1.000–1.010 are loanwords, and 240–250 are formed in Romanian by derivation or composition. The most important is, of course, a stratum of 81 loanwords and 102 Romanian creations belonging to the basic vocabulary of modern Romanian. This stratum comprises an even more valuable layer, that of the basic vocabulary, including 39 loanwords (cafea, cafenea, capac, caraghios, catifea, cântar, cearşaf, chef, chel, chibrit, chior, cioban, ciorap, cutie, dulap, duşman, farfurie, geam, geantă, habar, hai(de), haz, murdar, musafir, odaie, palavragiu, papuc, para, perdea, raft, sâc, sârmă, sobă, soi, tacâm, tavan, tavă, turc, tutun)and 67 terms formed in Romanian (e.g. degeaba, pălăvrăgi, sâcâi, sufragerie), with a core of six terms having the highest values and belonging to the principal lexical fund of today’s Romanian, namely cafea ‘coffee’, chef ‘1. desire, caprice; 2.feast’, duşman ‘enemy’, geam ‘pane, glass’, hai(de) ‘come on!; come along!; let’s (go)!’, and murdar ‘dirty’. At their turn, these most important loanwords have developed 18 new Romanian formations (e.g. duşmăni, gemuleţ, murdărie). Of some importance are also the 485–520 words (370–400 borrowings and 115–120 formed in Romanian) belonging to the outer layer of the bulk of the vocabulary, which includes literary and popular, occasionally and rarely used terms. The rest of the Turkish loanwords revolve with centrifugal tendencies at the periphery of the vocabulary (regionalisms, slang words, historical terms).

7. Those particularly interested in this topic may consult the second volume of our above-mentioned work, Dictionary of the Romanian Words of Turkish Origin, a historical and etymological dictionary of the 2.760 Turkish loanwords, presented with lexicographic means, including the first dates and attestations of their different meanings and formal variants borrowed or created in Romanian, as well as of their Romanian derivatives and compounds. The etymological explanations include the Turkish etymons and their formal or semantic alterations, as well as their correspondents in other Balkan languages. The abbreviation system, comprehensible for anyone and the French translation of the meanings allow the dictionary to be used also by researchers and public who do not speak Romanian.