AZERBAIJAN ix. Iranian Elements in Azeri Turkish

perhaps after Uzbek, the Turkic language upon which Iranian has exerted the strongest impact—mainly in phonology, syntax and vocabulary, less in morphology.



ix. Iranian Elements in Azeri Turkish

Azeri is, perhaps after Uzbek, the Turkic language upon which Iranian has exerted the strongest impact—mainly in phonology, syntax and vocabulary, less in morphology. Much of the Iranian interference is also present, albeit less strongly in other Turkic languages, e.g., Ottoman Turkish, but many features are specific to Azeri. The strong Iranian influence upon Oghuz Turkic began already in Central Asia. Since Persian Azerbaijan had been Iranian-speaking long before the Turkic immigration, there has been a thorough sub- and adstrative Iranian impact upon dialects of the area: some of them, such as Aynallū and Qašqāʾī (though these are sometimes classified as non-Azeri idioms), are particularly strongly Iranized. Furthermore, Persian as the culturally dominant language played a superstrative—or “roofing”—role which is obvious still today in southern Azerbaijan with its lack of linguistic standardization and long-standing general bilingualism.

Though generally recognized the Iranian influence on Azeri has not yet been investigated. For proper research in this field more information is required not only about different variants of Azeri but also about the local Iranian dialects of the contact regions, such as Tati and Ṭāleši.

Phonology. There is considerable interference at the phonological level. For example, all Azeri dialects spoken in Iran display phonotactic perturbations, partly due to Iranian influence. Especially affected is the Turkic sound harmony, although less than in Uzbek. We find, e.g., non-harmonic, i.e., invariable suffixes like -max (infinitive suffix): bil-max “to know,” etc. This and other similar phenomena are usually explained as results of Iranization, i.e., a breakdown of vowel harmony and a tendency to neutralize vowels and pronounce them centrally. However, not all deviations from the vowel harmony rules of Standard Turkish can be attributed to external factors.

Another exception to a common Turkic phonotactic tendency, heard mainly in educated speech, is the simplification of consonant clusters. Thus, [fikr] “thought” is heard instead of the integrated form [fikir], etc.

In phonetics there are several examples of presumably Iranian influence:

Vocalism. A tendency (also known in “Iranized” Turkic dialects of Central Asia) towards a fronted pronunciation of vowels, e.g., the shift of a > ä (K. Foy, 1903, p. 185). Accordingly, the short a in Arabic and Persian loanwords is rendered as a front vowel more often than in Turkish even in the neighborhood of emphatic and dorso-velar consonants, e.g., bäxt “happiness” (Turkish baht). The tendency may be valid also for the unrounded high vowels. Even if, as in Uzbek, New Uigur, etc., the phonetic distance between the front i and the back ï has been diminished, the phonological opposition is still maintained. (Gagauz and Karaim similarly have a fronted pronunciation of Common Turkic ï under Slavic influence.) In several Azeri dialects, ï is pronounced with the tongue slightly more advanced than in Turkish. However, this phenomenon is difficult to diagnose in terms of interference, since it is observed also in Kipchak languages (Kazakh, Karakalpak, and Tatar). There are also reasons to suppose the existence of a neutral schwa in some Azeri dialects; M. V. Monteil mentions a sound “à peu près l’e fermé persan,” e.g., mäne “me,” ade “his name” (instead of mäni, adï; cf. L. Johanson, 1978-79).

The tendency towards front pronunciation is also manifest in the shift of - > i- in words like ilan “snake” (Turkish yïlan). The southern dialects show strongly palatalized forms of k, g and l, e.g., [ćöć] for Common Turkic kök “root.” Thus, in spite of “disturbances” of the vowel system, the basic syllabic palatal correlation is maintained also in Azeri dialects.

The relatively open pronunciation of ä approaches that of Persian a. According to L. Ligeti’s observations, “cette voyelle azéri cherche toujours à se conformer, du moins en Perse et en Afghanistan, à la prononciation de l’a persan (ou tadjik) local” (1957, p. 114). In some Azeri dialects, as in Uzbek, the long a: is more or less rounded, e.g., [yå:d] “memory.” Typically Turkic vowels alien to Persian are sometimes replaced by more familiar sounds, i.e., ö > o, ü > u (M. Širäliyev, 1968, pp. 43f.), in non-initial syllables ï > u, etc. It is still an open question whether in some dialects remnants of Turkic vowel quantity oppositions have been preserved under the influence of the Persian long vowels. Whereas Turkish shows an aversion against Persian ow (< aw), this diphthong occurs frequently in Azeri (as well as Turkmen) dialects. Parallel to the Persian development ow < aw, even in native Turkic words low vowels are labialized in front of a labial element: aw > ow (often > o:), e.g., dowšan or do:šan “rabbit” (<*taβïšγa:n, cf. Turkish tavšan).

Consonantism. An Iranian feature in the consonantism is the palatalization of certain consonants (notably k, g, l) mentioned above. As in Turkman and many Anatolian dialects, there is a regular substitution of the un-Persian fortis q: initially by voicing, e.g., gal- “to remain,” guš “bird,” non-initially usually by spirantization, e.g., yaxïn “near,” yatax “bed.” (In Standard Azeri the latter change is restricted to the first syllable boundary.) Azeri, as Persian, but contrary to modern Turkish, has both [x] and [h]. There is no resistance to z (with substitution by g, as in some other Turkic languages). Initial fricative γ-, which does not occur in Turkish, is accepted to some extent, essentially in educated speech: γeyrät “zeal” (cf. Turkish gayrät), etc. The un-Turkic sound f is not only accepted in loanwords (häftä “week” against Kazakh apta), but also replaces native p in some dialects: e.g., if “thread.” (This phonetical development is also met with in Turkmen and Uzbek dialects.)

Suprasegmental features. In some dialects a special intonation pattern at the end of yes/no-questions, possibly due to Iranian influence, replaces the interrogative particle mI.

Morphology. Iranian derivational suffixes found in Azeri are -baz, -dan, -dar, -i, -kar, -keš, -päräst, -stan, and -xana, etc. Iranian prefixes such as bi- ~ be- and na-still play a role in word-formation. (All these elements were frequent in Ottoman Turkish but have now been largely abandoned.) The copula of the first person singular -(y)Am is usually explained as influenced by the corresponding Persian personal ending -am. Optatives like al-a-m “I (may) take” resemble in their structure the Persian subjunctive (present stem + personal endings, e.g. bar-am “I take”). The Persian perfect of the type āmada (ast) “he has come” (past participle [+ copula]) may have corroborated the use of -(U)b(dUr) as the usual perfect form among the Turks of Iran. (For the various perfect forms, see M. Širäliyev, 1967, pp. 213-20.) Sonqori, Aynallū and Qašqāʾī (possibly not classifiable as Azeri dialects) use Persian -tar as a comparative suffix, e.g., Aynallū yektär “better.” Iranian may also have influenced the aspect and temporal values, notably of the perfect forms, which function very much like the Persian perfect tense, e.g., yazmïšam “I have written.” gälibsän “you have arrived.” (For Azeri and Turkish -mIš, see L. Johanson, Aspect im Türkischen, Mainz, 1971, pp. 289f.).

Syntax. The impact of Iranian on Azeri syntax is particularly clear in the structure of complex sentences, especially in the sociolects of the educated. (Note that most of the features concerned occurred more frequently in Ottoman but have been given up in modern Standard Turkish; some subsist as substandard varieties.) There is a sort of replica syntax: Imitations of Indo-European language-type subordinative constructions are used instead of Turkic, left-branching, constructions, in which the subordinated elements are more or less expanded sentence constituents, morphologically based on verbal nouns, participles, and gerunds, cf. Bilirsänmi män kimäm? “Do you know who I am?” (instead of Mänim kim olduγumu bilirsänmi?); Heč kim dinmirdi, ondan ötrü ki, hamïnïn bu išdän xäbäri var idi (Mirzä Ibrahimov) “No one said anything, because everyone knew about this affair” (instead of Hamïnïn bu išdän xäbär olduγu üčün heč kim dinmirdi). As in Ottoman and Chaghatay, Persian subordinative conjunctions, alien to Turkic sentence structure, are widely used, particularly ki, which appears as a connective device between sentences of different kinds, e.g., Görmüšäm ki, onlar xošbäxt olublar “I have seen that they have become happy” (instead of Onlarïn xošbäxt olduglarïnï görmüšäm); Bir ata ki, bu išlär ilä mäšγul olan, onun oγlu da išgüzar olar “Also the son of a father who occupies himself with these things becomes skilful” (instead of Bu išlär ilä mäšγul olan atanïn oγlu da išgüzar olar); Sizin väzifäniz budur ki, tä’lim veräsiz “It is your duty to teach” (instead of Sizin väzifeniz, tä’lim vermäkdir); Män istärdim ki, sän gälsän “I would like you to come” (instead of Sänin gälmäyini istärdim); Atasï ona pul verir ki, o gedä bilsin “His father gives him money in order that he may be able to go” (instead of O, gedä bilsin deyä, atasï ona pul verir); Kitabï ačïrdïm ki, gapï döyüldü “I was just opening the book as there was a knock at the door” (instead of Kitabï ačdïgda gapï döyüldü). Like the Iranian subjunctive the optative is often used as a sort of subordinative mood.

Several conjunctions (and/or connective adverbs) of Persian origin are used even in Standard Azeri, e.g., ägär “if,” čünki “for,” gah . . . gah “now . . . now,” häm “also,” hämčinin “also,” häṛčänd “although,” härgah “if,” nä . . . nä “neither . . . nor,” näinki “not only,” yainki “or,” yaxud “or,” zira “for.” However, both these and conjunctions of Arabic origin occur frequently only in educated speech. Other frequent adverbs, modal words, and particles are bäli “yes,” bälkä “perhaps,” bäs “well,” “yes,” hämišä “always,” mägär “really,” etc.

Lexicon. The Iranian elements in Azeri are especially numerous at the lexical level. Azeri possesses a large number of Iranian loanwords missing or rarely used in Turkish (asan “easy,” bar “fruit,” javan “young,” čäp “crooked,” girdä “round,” huš “consciousness,” kar “deaf,” köhnä “old,” küčä “street,” mis “copper,” payïz “autumn, fall,” šänbä “Saturday,” turš “sour,” etc.); idioms, e.g., xahiš ediräm “please,” güzäšt elä “excuse me,” xudahafiz “good-bye”; numerous calques in phraseology (xoš gäl- “to please,” coined on Persian xoš āmadan, etc.); morphological contaminations as tanïš ol- “to know” = tanï-, cf. Persian dāneš “knowledge.” Some indefinite pronouns are of Persian origin, e.g., hämin “the same.” här “every,” här käs “everyone,” heč “any.” It must be left to further research to sort out the different layers of elements borrowed in the course of the long Irano-Turkic symbiosis. There may be some phonetic criteria, e.g., the majhūl vowel in forms like dost “friend” (Modern Persian dūst < dōst) points to an early date of borrowing, whereas ruzi “daily bread” (cf. Persian rūzī < rōzī) is a relatively late loanword.

The quantity of Iranian lexical elements differs significantly in the various forms of Azeri. Persian is more dominant in written than in spoken Azeri; the dominance is also more evident in the language of the educated. As for the innovatory vocabulary, northern Azeri often prefers Russian loanwords (e.g., vaγzal “railway station”), where southern Azeri chooses Persian ones or accepts European words through the intermediary of Persian or Turkish (e.g., istgah, gar, istasyon). Since, for several decades, there has been little, if any, cultural exchange between the two parts of Azerbaijan, the mutual intelligibility is decreasing. Whereas in Soviet Azerbaijan, the purist efforts have yielded considerable results, the Azerbaijani language of Iran, through school education and the growing influence of Persian mass media, remains very dependent upon Persian.



As G. Doerfer remarks in “Irano-Altaistica: Turkish and Mongolian Languages of Persia and Afghanistan” (in Th. A. Sebeok, ed., Current Trends in Linguistics VI, The Hague, 1970, pp. 217-34), there are “only sporadic remarks about the influence of Iranian on the Irano-Altaic languages.” (For details, see the bibliography added to Doerfer’s article.) A basic work on Azeri of Persian Azerbaijan is K. Foy: “Azerbajğanische Studien mit einer Charakteristik des Südtürkischen,” Mitteilungen des Seminars für Orientalische Sprachen zu Berlin, Westasiatische Studien 6, 1903, pp. 126-94; 7, 1904, pp. 197-265.

Important remarks are found in T. Kowalski, Sir Aurel Stein’s Sprachaufzeichnungen im Äinallu-Dialekt aus Südpersien, Kraków, 1939; L. Ligeti, “Sur la langue des Afchars d’Afghanistan,” Acta Orientalia Hungarica 7, 1957, pp. 110-56; V. Monteil, “Sur le dialecte turc de l’Azerbâydjân iranien,” JA 244, 1956, pp. 1-77.

General problems of Iranian influence are treated in K. H. Menges, “Indo-European Influences on Ural-Altaic Languages,” Word 1, 1933, pp. 188-93.

See also L. Johanson: “Die westoghusische Labialharmonie,” Orientalia Suecana 27-28, 1978-79, pp. 63-107 and “Reproduktion, Widerstand und Anpassung: Zur lautlichen Iranisierung im Türkischen,” in R. Schmitt and P. O. Skjærvø, eds., Studia Grammatica Iranica. Festschrift für Helmut Humbach, Munich, 1986, pp. 185-201.

Y. Z. Širvani, Äräb vä fars sözläri lüγäti, Baku, 1967.

A standard work on Azerbaijani dialectology is M. Širäliyev, Azärbayjan dialektologiyasïnïn äsaslarï, Baku, 1967.

See also G. Windfuhr, Persian Grammar, The Hague, Paris, and New York, 1979, pp. 188-89.

(L. Johanson)

Originally Published: December 15, 1988

Last Updated: August 18, 2011

This article is available in print.
Vol. III, Fasc. 3, pp. 248-251