CENTRAL ASIA xiv. Turkish-Iranian Language Contacts
xiv. Turkish-Iranian Language Contacts
I. Turkish peoples and languages in Central Asia.
Three Turkish languages came together in Central Asia, the territory covered by the modern Turkmen, Uzbek, Kazakh, Kirghiz, and Tajik SSRs, excluding Chinese Turkestan: 1. the Uighur or Eastern Turks (e.g., Chaghatay, later Uzbek), 2. the Oghuz, speaking Khorasani Turkish, 3. and the Kipchaks (Qipcaq/Qepčāq, e.g., Kazakh). The Turks first appeared in the 6th century under the Old Turkish empire. After the decline of the empire strong Turkish elements still remained north of the Aral lake (Oghuz and Kipchak), waging continuous wars against the caliphate and the Iranian dynasties. The Eastern Turks began entering Central Asia in the 3rd/9th century, but this was a gradual process (cf. Spuler, 1966).
In the 4th/10th century Säḷčük was converted to Islam, separated his tribes from the empire of the Oghuz Yabghu, and started the migration to the south. His successors conquered Nīšāpūr in 429/1038, and after the battle of Malāzgerd (463/1071) the Oghuz were able to penetrate vast parts of Anatolia.
Three Oghuz nations were formed gradually: 1. The Turkmen were the descendants of those Oghuz who, for some centuries, remained in the old Oghuz territories north of the Aral lake. 2. The Khorasani Turks were the descendants of those Oghuz who had remained in Khorasan (in the broad sense: from northeastern Persia to the Amu Darya basin); during the 10th-12th/16th-18th centuries the area inhabited by speakers of Khorasani Turkish was bisected by a wedge of Turkmen-speaking nomads migrating east, so that Khorasani Turkish is now spoken in the northeastern part of Persia and along the Amu Darya. 3. The western Oghuz (the differentiation into Osmanli Turks and Azerbaijanis only began in the Safavid period; cf. Doerfer, 1977, pp. 191-93).
The Kipchak tribes were held back in the north by the Oghuz and eastern Turkish dynasties for a long time, but in 807-08/1405 the Uzbeks (named after Özbeg, an outstanding ruler of the Golden horde, who at this time still spoke Kipchak, pushed forward to Ḵᵛārazm, Bukhara, and Astarābād and finally conquered the territory of modern Uzbekistan.
The Kazakh, however, revolting against the Uzbek khans, remained a free people in the 9th/15th century, although their political organization was chaotic (cf. Doerfer, III, pp. 461-68).
Modern Kipchak-Uzbek is the genuine descendant of the language of the Uzbek conquerors, whereas the language of the bulk of the modern Uzbek population is descended from an old Uighur dialect but contains a small admixture of Kipchak elements.
Western Uighur developed in three stages: 1. Ḵᵛārazm Turkish or Early Chaghatay (7th-8th/13th-14th cents.), 2. Classical and Late Chaghatay (9th-13th/15th-19th cents., a very late stage of Chaghatay, Sartish, was spoken in the 13th/19th and at the beginning of the 14th/20th cent.), and 3. Uzbek (after the Soviet takeover).
These three groups converged in the Amu Darya basin and have intermingled for centuries. Thus, the Uighur elements in many Khorasani Turkish dialects, for example, are as numerous as those of French in English, and Chaghatay poets could freely use Oghuz forms for metrical and other reasons. But in general the influence of Chaghatay was the strongest, and Chaghatay remained the only written language in the area until the 11th/17th century, being regarded as the most noble and original of Turkish languages (see, e.g., Ebn Mohannā, p. 73). However, despite its prestige it was not able to absorb the other Turkish languages.
Thus we may distinguish the following groups: 1. those of Uighur descent: Uzbek and modern Uighur in eastern Central Asia; 2. those of Oghuz descent: Turkmen and Khorasani Turkish (with Oghuz-Uzbek); and 3. those of Kipchak descent: Kazakh (with Karakalpak), Kipchak-Uzbek, Kirgiz, and (secondarily) Crimean Tatar.
Modern times. The Turkish languages in modern Central Asia clearly reflect the cultural and linguistic influence of Persia; although the amount of influence differs markedly in the individual Turkish languages (for a linguistic survey of the Central Asian Soviet republics see Tolstova, I, maps pp. 8-9, 164-65, 528-29; II, pp. 8-9, 160-61 and conclusion; cf. also the map in Philologiae Turcicae Fundamenta I). Following is a brief survey of the Turkish peoples of modern Central Asia and their languages. (It may be pointed out here that, generally speaking, the percentage of indigenous populations in the Central Asian republics is now increasing, while that of the Russians is diminishing.)
Turkmen and Khorasani Turks. In 1979, according to the census of that year, 1,892,000 Turkmen were settled in the Turkmen SSR, 92,000 in the Uzbek SSR, and 14,000 in the Tajik SSR (the figure 2,028,000 is given for the whole of the Soviet Union). In these figures is included an indeterminable number of Khorasani Turks (perhaps ca. 400,000; Doerfer, 1977) settled close to the border of Persia and along the Amu Darya (cf. Doerfer, 1987a, p. 429, and map in Doerfer, 1977, end); the majority of Khorasani Turks (ca. 1.5 mill.), however, live in northeastern Persia, some also in Afghanistan. Both Turkmen and Khorasani Turkish (a separate Turkish language [cf. Ivanow, 1926, p. 154], structurally situated between Azerbaijani and Turkmen, but closer to Azerbaijani [Doerfer, 1977]) belong to the Oghuz group of Turkish languages.
Oghuz-Uzbeks. These live somewhat to the northwest of Uzbekistan, along the Amu Darya (see map in Abdullaev, 1967). They may number several hundred thousands, though in the Soviet censuses they are included with the Uzbeks. Their language is closely related to Khorasani Turkish, belonging basically to the Oghuz group, with some admixture of Uzbek elements.
Kipchak-Uzbeks. Most of the Kipchak-Uzbeks live in approximately the same area as the Oghuz-Uzbeks (see map in Abdullaev, 1967); others are scattered throughout the republic (cf. the map in Philologiae Turcicae Fundamenta I). These are also classified as Uzbek in the censuses, but their number is presumably lower than that of the “Oghuz-Uzbeks.” The language of the Kipchak-Uzbeks is closely related to Kazakh. It is descended from the old Kipchak language of the 10th/16th-century Uzbek conquerors.
Uzbeks. The Uzbeks constitute the most numerous Turkish nation of Central Asia. Numbering a total of 12,456,000 in the entire Soviet Union according to the censuses (though some 100,000 of these are Oghuz- and Kipchak-Uzbeks), they are distributed throughout the Uzbek SSR (10,569,000), the Kazakh SSR (263,000), the Tajik SSR (837,000), the Kirghiz SSR (426,000), and the Turkmen SSR (243,000; this number probably includes some Oghuz-Uzbeks). The actual total today, however, is most probably about 17 million, as their average increase of population is enormous (according to Itogi, in 1970 they were only 9,195,000). The language of the Uzbeks is Uighur, with an admixture of Kipchak).
Karakalpaks. The Karakalpaks live south of the Aral Sea, in the Amu Darya delta and its surroundings. They number 298,000. Though they speak a Kazakh, or Kipchak, dialect, their culture and traditions differ considerably from those of the Kazakh nomads. Accordingly they have been regarded as a separate nation from early times.
Kazakhs. The Kazakhs are a relatively numerous Turkish nation. Of a total of 6,556,000 in the Soviet Union, 5,289,000 live in the Kazakh SSR, 620,000 in the Uzbek SSR, 27,000 in the Kirghiz SSR, and 80,000 in the Turkmen SSR. For a long time, however, there have been more Russians (40.8%) living in the Kazakh SSR than Kazakh (36%). The language is Kipchak.
Kirghiz. The Kirghiz population in the Soviet Union totals 1,906,000, with 1,687,000 in the Kirghiz SSR; 142,000 in the Uzbek SSR, and 48,000 in the Tajik SSR. Their language is Kipchak, representing a certain transition to Altay Turkish and, consequently, to the Turkish languages of southern Siberia.
Modern Uighur (formerly Taraṇči). 211,000 Uighurs are estimated to live in the Soviet union, with 148,000 in the Kazakh SSR and 30,000 in the Kirghiz SSR. There are also some in Afghanistan. The modern Uighur speak a language close to Uzbek.
Tatars. Finally mention should be made of the Crimean Tatars in Central Asia. In Soviet statistics they appear in the general category of “Tatary” together with the Kazan Tatars. However, the languages of these two groups are very different. The Crimean Tatars are immigrants to Central Asia who were expelled from the Crimean peninsula after World War II when they were accused of collaboration with the German army. There are 649,000 in the Uzbek SSR, 313,000 in the Kazakh SSR, 72,000 in the Kirghiz SSR, 80,000 in the Tajik SSR, and 40,000 in the Turkmen SSR, with a total of 1,154,000 (the figure of Kazan Tatars to be subtracted is unknown, but it appears to be small). Very little is known about their language today.
Before the October Revolution the Turkmen, Kipchak-Uzbeks, Kazakhs, and Kirghiz lived primarily as nomads, while the others were settled and had developed rural and urban cultures.
II. Language contacts.
Medieval times. The mutual influences between Iranian and Turkish languages in Central Asia began in Ḵᵛārazm rather early (see above). Ḵᵛārazm then became increasingly Turkicized, particularly after 436/1044, when it came under the Saljuqs (Tolstow, pp. 290-92). At the same time, however, Persian language was asserting itself in the same area (Spuler, 1966, p. 171), although the indigenous (Middle) Iranian languages Sogdian (q.v.) and Choresmian (q.v.) were also still spoken (Henning, “Mitteliranisch,” pp. 56-58, 84; see also chinese turkestan).
It is difficult to judge the influence of Central Asian Turkish in Choresmian, as the material at our disposal is very limited and hardly contains any Turkish elements. We cannot exclude the possibility, however, that the Choresmian translators deliberately did not employ any of the Turkish words that may have been in common use, in the same way that Kāšḡarī ignored the numerous Iranian loanwords found in other Turkish works of his time (5th/11th century; cf. Mansuroğlu, p. 108). The Choresmian interlinear glosses in Zamaḵšarī’s Moqaddemat al-adab (6th/12th cent.; ed. Togan) include a few Turkish words, for example, kyš = kēš “quiver” (Benzing, p. 37; Togan, p. 43), a word also attested in a Turkish-Khotanese word list (Bailey, p. 291), in the form kyeśä (i.e., kēš, not käš), but the Qonyat al-monya of Moktar Zāhedī Ḡazmīnī (d. 1257) contains only the Turkish phrase är oγlum “my son” in a Choresmian sentence (MacKenzie, p. 58, no. 123). A few terms relating to irrigation (arna “large canal” and yab “small canal”), which survived only at Ḵīva and among the Turkmen were supposed by Barthold (1956, p. 15) to be of Choresmian origin.
With the gradual disappearance of these Middle Iranian languages and the increasing spread of Persian and Turkish, with neither of them gaining the upper hand, an exchange of linguistic elements took place. In the Persian-Turkish symbiosis the Persians were at first the giving part and the Turks the receiving one. This process is strikingly illustrated by the special transformation of the ʿarūż metrical system as it went from the Arabs via the Persians to the Turks. Soon, however, the Turkish languages began to assert themselves, and, parallel to the Old Ottoman literature in the west (from the 7th/13th cent.) and the Kipchak literature of the Golden Horde (from the 8th/14th cent.), a remarkable literature in a language based on Uighur began to unfold in Central Asia (see chaghatay): under the Qarakhanids (5th-6th/11th-12th cents.), in Early Chaghatay (“Choresmian Turkish,” 8th/14th cent.), and in Classical Chaghatay (9th-13th/15th-19th cents.). Literary documents in the other Turkish languages spoken in Central Asia did not appear until much later (Turkmen, 12th/18th cent.: Philologiae Turcicae Fundamenta II, p. 725; Kazakh, late 12th/18th cent.: ibid., p. 743; Kirghiz, beg. of 14th/20th century, ibid., p. 760). The correspondence of the Qazaq (Kazakh) khans was also written in Chaghatay, and the languages of the older Turkmen poets was really a mixture of Turkmen, Khorasani Turkish, and Late Chaghatay.
Chaghatay and especially Uzbek (Özbäg) were mixed languages with an Uighur core and Kipchak elements (e.g., the Kipchak suffixes -li “provided with” and the ablative ending -dan). From the 10th/16th century onward we find both Uighur yaγï “enemy” (Pers. yāḡī “rebel”) and Kipchak ǰaw used in Chaghatay beside the older or hybrid form yaw (cf. Doerfer, IV, pp. 99-102). Chaghatay poetry also contains some Western Turkish/Oghuz elements (presumably Khorasani Turkish), used mainly for metrical reasons (e.g., Navāʾī, 9th/15th century, uses Oghuz yïγmïšam “I gathered” = long-short-long, instead of Chaghatay yïγmïš men = long-long-long, cf. Philologiae Turcicae Fundamenta II, p. 334); Gadāʾī, however, employs Oghuz expressions at will, not only for metrical reasons (cf. Gadāʾī, p. 5).
On the other hand, Persian poetry contains very few Turkish elements, aside from a few words that may be considered fads, although Persian historiography (e.g., Moʿīn-al-Dīn Naṭanzī) is replete with Turkish words (see below). However, the early Turkish elements in Persian usually disappeared again, and Persian grammar was not influenced at all. The comprehensive discussion of the Turkish elements in Persian and Persian/Tajiki in Doerfer (see IV, pp. 496-519, index) shows that the Turkish loanwords mainly belong to the spheres of administration and warfare.
Some Turkish loanwords had been adopted into literary Persian in the Saljuq period, some of which, for instance, qorbān “bow-case” (Oghuz, Kipchak, cf. Kāšḡarī qurmān), even found their way into Ferdowsī’s Šāh-nāma (e.g., Moscow, III, p. 88 v. 1358), and some more were borrowed in the Safavid period. These Safavid loanwords (of Azerbaijani origin) appear only in standard Persian, not in Tajiki Persian. In Central Asia, however, Chaghatay was the source of Turkish loanwords in Persian. At the same time it also served as the transmitter of Mongolian loanwords into Persian, for instance words such as qošiγun “vanguard,” later “brigade,” attested in Persian from Tārīḵ-eWaṣṣāf (comp. 728/1328) as qōšūn, qošūn and in Chaghatay from Qoṭb, the Turkish translation of Neẓāmī’s Ḵosrow o Šīrīn (comp. 743/1342; see Doerfer, I, pp. 406-10). A small number of loanwords antedating the Mongol invasion (ibid., I, pp. 39-41) are also mainly from the spheres of state administration and warfare.
The increase in the number of Persian loanwords in Chaghatay from the 5th/11th century was dramatic: Qarakhanid texts of the 5th/11th century contain only about 2% Persian loanwords, Early Chaghatay already has 22%, and more than half (57-63%) of the vocabulary of Classical Chaghatay was Persian (including Arabo-Persian). These Persian words were gradually completely assimilated and are still part of the Turkish languages of Central Asia. For instance, all the words in Navāʾī’s verse farqi-dīn tā qadäm füsūn u firīb “from top to toe enchantment and deceit” are used today in Uzbek. Chaghatay also borrowed grammatical elements from Persian, for instance, prepositions, conjunctions, eżāfa, yā-ye ešārat, and yā-ye waḥdat, elements quite alien to Turkish grammatical structure. This led to a certain revolution in the structure of these languages, which did not originally use prepositions; an expression such as tā qiyāmät küni “until the day of resurrection” is quite un-Turkish (cf. modern Turkish kıyamet gününe değin/kadar) but still used in Uzbek.
Modern times. The most obvious intermingling of peoples and languages in Central Asia is that seen in the Uzbek SSR between the Uzbeks and the Tajiks, and it is the interaction between Uzbek and some other Turkish languages in this area and Tajiki Persian that has had the strongest impact on the linguistic history of Central Asia. On the whole, however, the influence by Tajiki Persian is strongest in Uzbek, less strong in Oghuz-Uzbek, and only very slight in Kazakh and Kirghiz. Generally speaking, the Persian loanwords in Uzbek are typically Tajiki, while those in the Turkmen and Khorasani Turkish show standard Persian characteristics (presumably of the Khorasani variety). The loanwords also come from different semantic spheres (cf. Saparov, passim, for Turkmen and Uzbek-Oghuz). The Tajiki literary language has been strongly influenced by Uzbek, though not as strongly as Uzbek by Tajiki. Only in certain Tajiki dialects is the Uzbek influence stronger even than the Tajiki influence in Uzbek. Yaghnobi and the Pamir languages contain both Tajiki and Uzbek elements, but there the Uzbek elements have probably been transmitted via Tajiki rather than borrowed directly.
Persian (Tajiki) influence on Turkish languages. The phonology of modern Uzbek is identical with that of Tajiki Persian (which differs from that of standard Persian). The original Turkish vowel phonemes /ä, ö, ü, ï/ in first syllables have become /e, o, u, i/; in other than first syllables we find /a/ instead of /e/; the development of Turkish *a (in first syllables) is twofold (cf. Nauta): partly > [å] (written o), a very low back rounded vowel, partly > [æ] (very low front unrounded); Turkish *o, which was most probably pronounced low back rounded, has become low-mid back rounded, written ŭ, as in Tajiki. Only the Uzbek dialects to the north of Tashkent (roughly those in the Kazakh SSR) have retained the original Turkish vocalism (Fazylov and Chichulina, pp. 65-78), and variations are also found elsewhere. Tajiki has preserved q- (originally foreign to Iranian languages, but found in Arabic and Turkish) under the influence of Uzbek.
Tajiki Persian prepositions (e.g., to “until”) and conjunctions (e.g., čunki “because”) are commonly used. In word formation a number of prefixes and suffixes (cf. Borovkov, pp. 716-27; Fazylov and Chichulina, pp. 131-33) have become productive (e.g., kam- “little,” no- “non-,” -jon “dear”); examples: kam-suv “insufficiently irrigated” (cf. Taj. kam-ob, Pers. kam-āb), no-tŭγri “incorrect” (cf. Pers. nā-dorost; this word was borrowed in Karakalpak as natuwrï; and back in Tajiki as notŭḡrī), otajon “father dear” (cf. Pers. bābā jān).
The ratio of Tajiki Persian words in Uzbek was about 2:3 in the 19th century. However, according to a study by Fazylov and Chichulina (p. 147), in the language of the Uzbek press the Tajik vocabulary has gradually diminished from 45% to 25%, while the Russian vocabulary has increased, from 2% to 15% over a 13-year period from 1927 to 1940. The percentage of Persian (and Arabic) words in Uzbek has always been greater than in Kazakh, and, according to a statistic taken in 1982 (Rüstemov, p. 21), the ratio of Turkish, Persian, and Russian words in an Uzbek text translated from Russian was 56:31:13, whereas in the Kazakh translation it was 72:15:13.
To evaluate the number of non-Turkish words in Turkish languages one may employ a “p-/f-test.” This is a test based on the fact that in Central Asian Turkish languages f- is not found initially in words of Turkish origin and p- only in a few words where it has resulted from assimilation (e.g., Turkman pïčaq “knife” < bïčaq). Aside from words of Russian origin, words with initial f- are therefore always of Persian (or Arabo-Persian) origin, and words with p- are almost always of Tajiki Persian origin (some rare and colloquial words have secondary p- < f-, e.g., palakat “bad luck” < Pers. falākat, Taj. falokat; palon “N.N.” < Pers. folān, Taj. falon). Examples: padar “father,” padarkuš “patricide,” pazanda “chef de cuisine, master cook”; favvora “fountain” (Pers. fawwāra), favquloda “extraordinary” (Pers. fawq-al-ʿāda), faje “tragic” (Pers. fajīʿ, Taj. fajēʾ).
Applying this test to words in alphabetical order from pa- to pal- and from fa- to fal- (including words in po- and fo- < pā- and fā-, but omitting words containing Turkish suffixes, e.g., Uzbek foydalan- “to use” < foida “use,” Pers. fāʾeda, Taj. foida) we find in Borovkov’s Uzbek dictionary (which contains a total of 40,000 words) 116 words (82 words with p-, 34 with f-), that is 0.29% of the entire dictionary.
In the other Turkish languages of Central Asia the number of p-/f- words is much smaller. In Turkmen, where f- has become p- (e.g., pāl “divination” < Pers. fāl) we find in Baskakov et al. (1968; 40,000 words) 46 words (35 with original p-, 10 with original f-) with pa- and pä- (= Uzbek pa-, cf. Turkm. päydā = Uzb. foida) that is 0.11%. The results for Uzbek-Oghuz (cf. Saparov) are similar.
For Karakalpak, which also has f- < p-, we find in Baskakov’s dictionary (30,000 words) 26 words (18 with original p-, 8 with original f-), that is 0.09%.
For Kazakh, with f- > p- (cf. payda “advantage”), only 10 words (5 each with original p- and f-), that is 0.06%, are found in Shnitnikov’s dictionary (about 17,000 words).
For Kirghiz, where f- and p- have become b- (except in some southern dialects; e.g., bal “divination,” bada “stable”), we find in Yudakhin’s dictionary (40,000 words) 25 words (20 with original p-, 5 with original f-), that is, 0.06%. Several of them are, however, attested only in the southern dialects, spoken in areas close to Uzbek and Tajik (e.g., payvant “graft”).
For Uighur we find in Kibirov and Tsunvazo’s dictionary (16,000 words) 13 words with p- (9 with original p-, 4 with original f-), that is, 0.08%. In the Xinjiang (Sinkiang) dialect the number of words beginning with f- and p- is substantially higher, indicating a stronger Tajiki influence. Jarring’s dictionary (about 8,000 words) contains 16 words with p-, 5 with f-, a total of 21 words = 0.26%.
These statistics are paralleled by a similar distribution of phonetic developments. Iranian words with consonants in the initial position that in the Turkish languages are confined to the internal position (e.g., z) are borrowed in the original Iranian form, for instance, Kirghiz, Kazakh, Karakalpak, Uighur zor, Uzbek zur, Turkman zōr “power.” On the other hand, Tajiki contains consonants, both initial and other, that are either not found in Turkish at all or disappeared early in the history of Turkish languages. Thus Persian h- (including Ar. ḥ-) has remained h- in some languages, probably where the original Turkish h- has also been preserved, at least in some instances (cf. Doerfer, 1981-82; e.g., ḥokm “decision, order” > Karak. hükim, Uzb. hukm, Uigh. höküm), and in Turkmen it hardened into x- (xöküm). But in Kazakh and Kirghiz, sometimes also in Karakalpak, it was lost (e.g., Karak. yesap, “count” < ḥesāb, Kaz. ükim, Kirgh. öküm). Internally h remains in Karakalpak, Uighur, and Uzbek (e.g., baha or baho “price” < Pers. bahā, Taj. baho), becomes Turkmen x (baxa), Kazakh -γ- (baγa), and is lost in Kirghiz (baa, pronounced bā). Persian ḵ and ḡ mostly remain but occasionally become q (e.g., Kirgh. qabar “information” < Pers. ḵabar; Uigh. miq, Kaz., Kirgh. mïq, Karak. mïyïq “nail” < Pers. mīḵ; Kaz. baq or baw, Kirgh. baq “garden” < Pers. bāḡ). Internal -v- remains in Uzbek (devor < Taj. devor, quvvat “strength” < Pers. qowwat), Uighur (quvvät), and Turkmen (dīvār, quvvat); becomes Karakalpak -w- (diywal, quwat), Kirghiz -b- (dubal, qubat); and is lost in Kazakh (dual, quat). Internal ʿayn is normally lost (e.g., Kirgh., Karak. saat, pronounced sāt, Uzb. soat, Uigh. saät “hour” < Pers. sāʿat, Taj. soat) but becomes -γ- in Kazakh (saγat) and Turkmen (sāγat).
The situation in modern Uighur is similar to that in Uzbek. It too has assimilated certain prepositions and conjunctions (e.g., ta “until,” čünki “because”), and forms such as natoḡra “incorrect”; however, it has preserved the “umlauted” vowel phonemes (only exception ï > i). The remaining Turkish dialects, although containing a certain amount of Persian/Tajiki vocabulary, have resisted the Iranian influence more successfully: p-, found in internal and final position in Turkish, has been widely substituted for initial f-, and in Kirghiz (in the central and northern dialects) normal Turkish b- has been substituted for p-. In the easternmost reaches of the Turkish language area, in south Siberia, Persian loanwords are found only quite sporadically, such as sabïn “soap” (Pers. ṣābūn), mal “cattle” (Pers. māl), ḵuday “God” (Pers. ḵodā(y); see Baskakov and Inkizhekova-Grekul).
In Kirghiz we find mainly loanwords designating religious and abstract concepts: ayïp “sin” (Pers. ʿayb), payda “profit” (Pers. fāʾeda); economic terms (commerce, agriculture): baa “price” (Pers. bahā), bul/pul “money” (Pers. pūl), baq “garden” (Pers. bāḡ), piyaz “onion” (Pers. pīāz), dan “grain” (Pers. dān), paxta “cotton” (Pers. paḵta); material culture: dubal “wall” (Pers. dīvāl); most of these are terms unknown to nomadic pastoral life. In a very few instances, however, Persian words have—for unclear reasons—replaced Turkish ones, for instance, Turkish soγan “onion” was replaced by pïyaz from Persian.
Some phonological phenomena show that the Iranian loanwords entered many of the Central Asian Turkish languages rather late, at various times, and sometimes perhaps also not directly. To see this we may examine words beginning with g- and š-. Initial g- is rarely preserved in the Central Asian Turkish languages, for example, gül “rose” (Uzb. gul). In most cases the development is twofold, for example, “grave” (Taj. gŭr, Pers. gūr, older gōr) is gūr in Uzbek, gör in Turkman (where, however, initial g- is usual), Karakalpak and Uighur, but kör in Kazakh (only some dialects close to Uzbekistan have gör, Kenesbaev, p. 49) and in Kirghiz (only southern Kirghiz, close to Tajikistan, has gör). In Uzbek, Uighur, and Karakalpak the initial g- has been preserved because of the strong influence of Tajiki (in Karakalpak presumably via Uzbek, as in the case of gör in Kazakh dialects). The development in Kazakh and Kirghiz, however, is the most genuine, as initial g- was not originally found in Turkish, and so it was changed to k-.
Another interesting phenomenon is the development of Persian (Tajiki) š-, which is retained in Uzbek, Turkmen, and New Uighur, as well as in Kirghiz (sorpo “broth” vs. šorpo, šorpa is an exception and perhaps a loanword from Kaz. sorpa). In Kazakh and, to a lesser degree, Karakalpak the situation is different, however. In original Turkish words these languages have š- > s-, for example, sol “that one” (< šo-ol). But Turkish š- did not become s- until the 12th/18th century (Doerfer, 1971, p. 330), and the development is not yet completed in Iranian loanwords, in which š- is still commonly found in southeastern Kazakh dialects (also for Turkish words) and in Karakalpak (both close to Uzbekistan), while northwestern Kazakh dialects mostly have s- (Qaliev and Sarïbaev, pp. 46-48, 123-28, 131-32). In Mongolian loanwords in the Kazakh and Karakalpak written languages š- became s-, as also in genuine Turkish words (sïltaw “cause,” sïba- “to smear”; these words entered Kazakh and Karakalpak during the Yuan/Elḵan period or only shortly afterwards, but not later than the 9th/15th century; Kirghiz still has šïltō, šïba-). Russian loanwords, on the other hand, always have š- (e.g., šar “globe”; the oldest of Russian loanwords entered Kazakh in the 13th/19th century, mostly in the Soviet time). Following are some examples of Tajiki Persian loanwords with original š-. There are three categories: (a) š- is retained, (b) vacillation between š- and s-, and (c) development of š- to s-:
(a) Šalbar “trousers” (Pers. šalvār), šam or šem “candle” (Pers. šamʿ), šaptalï “apricot” (Pers. šaftālū), šarap “wine” (Pers. šarāb), šat “glad” (Pers. šād), šen “rank” (Pers. šaʾn), šeker “sugar” (Pers. šakar), šek, šäk “doubt” (Pers. šakk), šäkirt “disciple” (Pers. šāgerd), šäher “town” (Pers. šahr). (Karakalpak always has forms in š-; for Kazakh Radloff has šalbar, šam, šat; Katarinskiĭ has šalbar, šam, šeker, šek; Bukin has šalbar, šam, šaptali, šakirt, šaxar, but seker “sugar”).
(b) Šaytan or saytan “devil” (Pers. šayṭān), šalï or salı “unhusked rice” (Pers. šālī), šart or sert “condition” (Pers. šarṭ), šembi or sembi, senbi “Saturday” (Pers. šanba). (Karakalpak has š- in all cases; for Kazakh Radloff gives šaytan or saytan, sert, sembi; Katarinskiĭ saytan, sembe; Bukin sert.)
(c) Samal “breeze” (Taj. šamol “wind,” Pers. šamāl “north [wind]”), semser “sword” (Pers. šamšēr), serik “companion” (Pers. šarīk), sorpa “broth” or šōrbā, sor “salt marsh” (Pers. šōr “salty”). (Karakalpak has šamal, in dialects also samal, semser, šerik, sorpa, sor; on the whole, Karakalpak has a preponderance of forms with š- and in this respect resembles the southeastern Kazakh dialects; for Kazakh Radloff has samal, simsär, serik, sorpa, sor; see also Menges, 1947, pp. 81-94.)
In conclusion, in Persian loanwords that entered Kazakh and Karakalpak before the 12th/18th century š- became s- as in genuine Turkish and Mongolian words. In Persian words that were borrowed at a later time, for example, from the 12th/18th century onwards (when Russian words also began to be borrowed), after the original sound change had taken place, an initial š- was either retained or changed to the more common s-.
Geographically this shows that most Central Asian Turkish languages got their loanwords, not from standard (western) Persian, but from Central Asian Persian (Tajiki), some also from Khorasani Persian. Thus, Khorasani Turkish šämāl “wind,” Turkmen šämāl “wind, breeze,” Uzbek šamol “wind,” Uighur and Karakalpak šamal “wind,” Kazakh samal “breeze,” Kirgiz šamal “strong wind” are not from standard Persian šemāl, šamāl “north, north wind” (Azerbaijani šimal, idem), but from Tajiki šamol “wind” (in general; cf. also Khorasani Pers. šumol “cool breeze,” Ivanow, 1928, p. 287). But Central Asian and Khorasani Persian have old geographic contacts with the Turkmen, Uzbek, and Uighur, while their contact with Kirghiz may be younger (the Kirghiz presumably entered their modern area no earlier than the 10th/16th century). Kazakh and Karakalpak, which do not have direct contact with Central Asian Persian, probably received the Persian loanwords via Uzbek, but this question still needs to be investigated.
Turkish influence on the Iranian languages of Central Asia. The most important influence of a Turkish on an Iranian language in Central Asia is that of Uzbek on Tajiki Persian, which has been only insignificantly influenced by other Turkish languages (Khasanov, pp. 194-96, mentions some Kazakhisms in a Tajik border dialect). The influence of Uzbek on Tajiki was acknowledged as early as 1861 by Grigor’ev, later also by Kuznetsov and others (Oranskiy, pp. 19-21, 24-25, 35-38, 43, 48-50, 54-63). In order to describe this influence we have to distinguish three principal varieties of Tajiki: the literary language, the northern dialects, and the southeastern dialects.
Literary Tajiki. The Uzbek loanwords in literary Tajiki (originally a language developed by the Soviets on the basis of dialects that differed most strongly from standard Persian, written in Arabic script until 1930, in Latin script from 1930-40, since then in Cyrillic) were described by Doerfer (based on Bertel’s; see, in particular, the index in Doerfer, IV, pp. 554-72). We must distinguish two layers of Turkish loanwords in literary Tajiki: On one hand there are loanwords that are also attested in literary standard Persian and therefore quite old, for instance, āzūq, āzūqa, Tajiki ozuqa “provisions,” or Persian yāl, Tajiki yol “horse’s mane” (both attested since the 5th/11th century, see Doerfer, II, pp. 56-57, and IV, pp. 105-06). Such words represent a common standard Persian and Tajiki Persian stock of loanwords. On the other hand, many Uzbek loanwords are only found in Tajiki Persian, for instance, ang “intelligence” (Bertel’s, p. 26; Doerfer, II, p. 130).
The number of Uzbek loanwords in literary Tajiki is quite high, exceeding by far the number of Turkish loanwords in standard Persian, and are common in the colloquial language, as well. It should also be remembered that of the some 2,000 Turkish words listed in Doerfer’s dictionary some were not integrated loanwords but only Turkish words used in particular contexts. Sometimes the Turkish words represented fads that were used for a short period only and then went out of common use. Certain authors, for instance Moʿīn-al-Dīn Naṭanzī (author of Montaḵab al-tawārīḵ-e moʿīnī, comp. 817/1414, ed. J. Aubin, Tehran, 1336 Š./1957), are also particularly fond of showing off their knowledge of Turkish or Mongolian. The total number of integrated Uzbek loanwords in literary Tajiki is therefore more likely to be only about a third of the 2,000 listed by Doerfer.
It is interesting to note that the Uzbek loanwords in Darī, or Afghan Persian, often show different forms from those in Tajiki and more often coincide with the standard Persian forms (cf. Kiseleva and Mikolaĭchik). Thus to Tajiki qapidan “to catch” corresponds Darī and literary Persian qāpīdan. The Uzbek loanwords in literary Persian, literary Tajiki, and literary Darī still need to be investigated; however, it appears that Darī contains fewer Turkish loanwords than Tajiki.
A comparison between the Turkish loanwords in standard Persian and Tajiki Persian further reveals that Persian contains a fairly large number of loanwords from the sphere of state and government, whereas Tajiki Persian has a larger number of loanwords from a lower social level, including words such as yarma “groats.” This would seem to indicate that Persian received the Turkish loanwords via the higher classes, rulers and officers, reigning nobility and officers, but Tajiki through its contact with the Uzbek common people.
The influence of Uzbek may also have contributed to the preservation of the initial voiceless velar plosive q- in Tajiki Persian, attested since the 10th/16th century, which has become a voiced velar plosive in standard Persian (written ḡ, which in Arabic, however, is a fricative), possibly under Azeri influence (cf. Oranskiy, pp. 59-60, 62-63, quoting Lazard).
Finally, it must be pointed out that the form of the Uzbek loanwords in Tajiki are much closer to Old Turkish than those of modern Uzbek. Thus, for instance, modern Uzbek botir “hero,” but Tajiki botur, Old Turkish bātur. Similarly both Persian qošūn, qūšūn “army” and Tajiki qŭšun are closer to Mongolian qošiγun than Uzbek qŭšin “army.”
Northern Tajiki dialects. The northern Tajiki dialects, that is, the dialects spoken in Ašt, Brič-Mulla, Kanibadam, (Kand-a Bādām), Kassansay, Rištan (Reštān), Čust, and Šaydan, and others, have been very strongly influenced by Uzbek and have even been characterized as Turkish languages in statu nascendi (thus Doerfer, 1967, p. 57; cf. Rastorgueva, 1963, 1964). Only the mountain Tajiks may have escaped this influence to some degree (Oranskiy, p. 83). Among the Turkish features in northern Tajiki some are common Turkish, others are typically Uzbek (cf. Doerfer, 1967, pp. 52-56). One of the most striking Uzbek constructions in northern Tajiki is the possessive construction muallim-a pisar-aš (lit., “teacher-genitive [< -rā] son-his”), which has replaced common Tajiki pisar-i muallim (Pers. pesar-e moʿallem) and which is identical with Uzbek muallim-ni ŭḡl-i, the common Turkish syntactic construction (cf. Oranskiy, p. 84). Identity between the genitive and accusative is found in both northern Tajiki and many Uzbek dialects (north. Taj. -a < *-rā, Uzb. -ni), compare northern Tajiki inkitob-a ḵondagiman “I have read this book,” Uzbek bu kotob-ni ŭqiyan-man (Rastorgueva, 1974, pp. 138-39). Many suffixes, even inflectional, have been borrowed, for instance, the ablative suffix -dan (yakŭm klasdan “from the first class,” literary and southern Taj. az klas-i yakŭm, Pers. az kelās-e yakom).
Northern Tajiki has borrowed a number of Turkish verbs not found in literary Tajiki and the southern dialects (e.g., ič-i-dan “to drink”; Doerfer, 1967, p. 19). The construction -miš + kardan (also found in the west, in Ṭāleši, Kurdish, Zāzā, Tāti, as well as several non-Iranian languages, see Doerfer, I, p. 33) is quite common, for example, bŭl-miš kardan “to divide” (Uzb. bŭl-; cf. Doerfer, 1967, pp. 67-70); sometimes the -miš form is replaced by the simple verbal root, for instance, tula kardan beside tula-miš kardan “to pay” (Uzb. tŭla-). The construction is an extension of the common method of forming “compound” verbs from Persian verbal nouns or Arabic nouns, for instance, pardāxt kardan “to pay,” fatḥ kardan “to conquer.” It must be kept in mind, however, that the suffix -miš, unlike in modern Azeri Turkish, is no longer common in modern Uzbek, where it is used only sporadically and in a few dialects. It must therefore have been adopted in northern Tajiki at the time of the old Eastern Turkish Chaghatay (Kononov, p. 272), where it was common as a verbal noun only in the 5th-8th/11th-14th centuries, but as a finite verb also in the 7th-8th/13th-15th centuries, though more often found in poetry than in prose (Eckmann, pp. 141, 167-69), for instance, bular ketmïš “they have gone.” The suffix -miš, it should be pointed out, is not, however, directly connected with the suffix -mīšī found in Persian between the 7th/13th and the outset of the 9th/15th centuries (cf. Doerfer, I, pp. 32-33).
Whereas northern Tajiki has borrowed even case endings (even -ga dative, -da locative, -dak equative, -gača terminal, cf. Doerfer, 1967, pp. 54, 62), the influence of the Uzbek verbal system is seen both in northern Tajiki and in the literary language (Doerfer, 1967, pp. 52-55; Bertel’s, pp. 560-63 ). One might say that, although the Tajiks speak Tajiki, they feel Uzbek, as it were.
There are numerous verbal constructions borrowed from Uzbek, of which the following from literary Tajik may serve as examples: navišta dodan < Uzb. yozib bermoq “to write for another’s profit,” literally “to give writing”; navišta giriftan < Uzb. yozib olmoq “to write for one’s own profit,” lit. “to take writing”; xŭrda didam < Uzb. yeb kŭrdim “I tried to eat, I tasted the meal,” lit. “I looked eating”; kanda karda firistod < Uzb. kulib yubordi “he burst out laughing (suddenly and surprisingly),” lit. “he sent laughing”; kitobro ḵonda šudam < Uzb. kitobni ŭqib bŭldim “I read the book through (completion of an action),” lit. “I became reading the book.” The syntactical use of participles, infinitives, and gerunds is just as in Uzbek (both in literary and northern Tajiki): kitobi man ḵondagi, kitobi ḵondagiam, kitobi ḵondagii man < Uzb. yozgan kitobim (also mening yozgan kitobim, men yozgan kitob “the book I’ve read,” literally “my-book reading, the book of my reading.” A kind of referative aspect is formed by the participle in -dagi/-tagi, for example, raftagi-am < Uzb. ketgan-dir-man “(one may suppose that) I have gone,” lit. “I am gone.” As in Uzbek the infinitive is used frequently, for example, (northern Tajik) kalkos-va daromadan bat < Uzb. kolḵozga kirmiš blan “after (with) having come to the kolkhoz,” lit. “to-the kolkhoz to come after (with).” The use of the perfect participle in -da/-ta as a gerund, also found in standard Persian, is common in long sentences, where Uzbek would use participles in -ib, for instance, piramard ba piyola reḵtani čoi nimsard šudai čoynikro lozim nadonista, onro bardošta az nŭlaš tamom nŭsid “the old man, not thinking it necessary to pour the tea, which had cooled down in the teapot, into the cup, raising (the teapot) drank (it) up from the spout.”
Southern Tajiki dialects. These dialects are spoken in the southernmost areas where Tajiki is spoken, along the Afghan border (cf. Rozenfel’d, p. 4: Karategin, Kulyab, Vaḵio, Darvāz, Vanj, Badaḵšān; Oranskiy, pp. 62-64, 65-71). Here the Uzbek influence is minimal and does not affect the grammar. Rozenfel’d lists only 21 Uzbek loanwords with initial q- in southern Tajiki of a total of 198 (= 10.6%), compared to 96 listed by Rastorgueva (1963) for northern Tajiki of a total of 134 (= 71.6%), and 52 by Bertel’s for the literary language of a total of 348 (= 14.9%). Among the Uzbek loanwords in the southern dialects are some not found in the literary language or the northern dialects (though perhaps only coincidentally), for instance, tul “propagation of cattle” (Uzb. tŭl) axsum “stubborn horse” (Uzb. axsum, aqsum), yalama “rice pap with meat” (Uzb. idem). It is also clear that in many cases the Uzbek dialect forms found in southern Tajiki differ from those found in the literary language and northern Tajiki, for instance, southern Tajiki qap “sack” (north. Taj. qop = liter. Uzb.). Very few southern Tajiki verbs contain Uzbek elements; one rare example is qapidan “to catch” (from Uzbek qap-; see also above).
Yaghnobi. Yaghnobi contains a fairly large number of Uzbek loanwords. Most of these, however, it shares with Tajiki (literary and/or dialects). Of the 139 loanwords found in the dictionary of Andreev and Peshchereva 123 also occur in literary Tajiki (e.g., boy “rich” = Uzbek); seven words occur in some dialect (e.g., tul, see above); and only nine are not attested in the Tajiki glossaries of Bertel’s, Rastorgueva, and Rozenfel’d (e.g., qušqoñ “bridle” < Uzb. quyušqon), which, together with the fact that the Yaghnobis are surrounded, not by Uzbek, but by Tajiks, renders it likely that the “Uzbek” loanwords of Yaghnobi have come via (southern) Tajiki dialects (cf. Oranskiy, p. 114).
Pamir languages. The Uzbek influence in the Pamir languages is also principally of lexical nature and has come via Southern Tajiki dialects. Most loanwords with initial y- and q- are from Uzbek but are found in all or most Tajiki, Yaghnobi, and Pamir dialects, for instance, yol “horse’s mane,” yoš “jung,” qoq “dry,” qoz “goose,” and others (cf. e.g., Sköld, pp. 132-68 nos. 52, 87, 112, 255a, 346-47, 354a, 355, 364, 380, 381, 383, 412, 417b, 425).
Central Asian Persians. There is no detailed description of the small group of Persians (about 30,000) living in the Turkmen SSR, above all around Marv (Mary), Bukhara, and Samarqand (cf. Oranskiy, pp. 124-25). According to oral information they understand standard Persian and Uzbek, but not Tajiki.
Baluch. About 7,800 Baluch live around Marv (cf. Sokolova, 1953, I, p. 77; Oranskiy, pp. 121-24; see also baluchistan iii, in EIr. III/6, esp. pp. 634-35). They are late immigrants who came from Persia and Afghanistan about 1900. In 1920-40 their second language was still Persian, which today, especially among the younger generation, has been replaced by Turkmen. Elfenbein’s word list contains just a few Turkish loanwords, not, however, of Turkmen origin, as might be expected, but from Persian or Darī Persian, for instance, dulma “forcemeat” (Pers. dolma < Azeri dolma, cf. Doerfer, III, pp. 203-04; eraγ “attire” < yarāq; γarāwūl “watchman” < qarāvul, see Doerfer, s.vv.; čol “desert” < Darī čol; γamčune “whiplash” < Darī qamčina). The only apparently Turkmen loanword is γum “sand” (Turkm. gum).
Kurdish. The Kurds of Afghanistan, numbering about 20,000 and settled around Ashkhabad, are Kurmanjī and immigrants from Khorasan (Bakaev, p. 6; cf. Oranskiy, pp. 153, 160; Sokolova, 1953, I, pp. 78-103). They also speak Turkmen. The Turkish loanwords registered by Bakaev proves that the Kurds of the Turkmen SSR must originally have come from Azerbaijan (or a region close to Azerbaijan). Thus we find (Bakaev p. 226) k’oti or k’uti “bad” (Ivanow, 1927, p. 229: kuti) from Azeri Turkish kötü (Khorasani Turkish uses pis and gändä—both from Persian—and yaman “bad,” the last also in Turkman); also forms as yoḵ “not existing,” batlaḡ “swamp” (Pers. bātlāq) are typically Azerbaijani. Like western Kurdish this dialect also contains mixed verbal constructions with -miš (e.g., bašlä-mis kïrïn “to begin”).
Arabs of Central Asia. The Arabic spoken by the Arabs of Central Asia, above all Bukhara, has been influenced by both Persian/Tajiki and Uzbek (Fischer; Vinnikov; Doerfer, 1969, with further bibliography). Particularly noteworthy are features such as the omission of the article and the position of the verb at the end of the sentence, as well as other features common to Tajiki and Uzbek. Persian features include the expression of the durative aspect with mi-, ki used as relative particle, eżāfa, etc. Uzbek features include the interrogative particle -mi, the position of adjectives before the substantives. In the numerals we find both the Uzbek construction ʿašer ḵams “15” (Uzbek ŭn beš) and the Persian construction ʿašrīn-u arbaʿ “24” (Tajiki bist u čor).
Summary. Central Asia, and particularly the territory inhabited by Uzbeks and Tajiks, is an area of extensive language mixture. The long symbiosis of these two nations has caused, not only in cultural but also in linguistic respect, what Bausani has called “an identical Muslim cultural world” (Pagliaro and Bausani, p. 752). Turkmen and Khorasani Turkish all contain a strong admixture of standard Persian elements borrowed from Khorasani Persian, while all the other, smaller, languages have been influenced by Uzbek and/or Tajiki Persian.
General: Barthold, Turkestan. Idem, Four Studies on the History of Central Asia I, tr. V. and T. Minorsky, Leiden, 1956.
G. Doerfer, Turkische and mongolische Elemente im Neupersischen, 4 vols., Wiesbaden, 1963-75.
Ebn Mohannā, Ḥelyat al-ensān wa ḥalbat al-lesān, ed. K. Refʿat, n.p. [Istanbul?], 1338-40/1919-22.
A. Herrmann, Atlas of China, Cambridge, Mass., 1935.
Handbuch der Orientalistik, ed. B. Spuler, Leiden and Cologne: I/IV: Iranistik 1, Linguistik, 1958; 4, Literatur, 1968; I/V: Altaistik 1, Turkologie, 1963; 5, Geschichte Mittelasiens, 1966.
Itogi vsesoyuznoĭ perepisi naseleniya 1970 goda IV: Natsional’nyĭ sostav naseleniya SSR, soyuznykh i avtonomykh respublik, kraev, oblasteĭ i natsional’nykh okrugov, Moscow, 1973.
L. Krader, Peoples of Central Asia, Uralic and Altaic Series 26, Bloomington, Ind., 1963.
K. Menges, “Altaic,” in EIr. I, pp. 909-12.
Naselenie SSR po dannym vsesoyuznoĭ perepisi naseleniya 1979 goda, Moscow, 1980.
B. Spuler, “Geschichte Mittelasiens seit dem Auftreten der Türken,” in HO, 1966, pp. 123-310.
S. P. Tolstova et al., Narody Sredneĭ Azii i Kazakhstana, 2 vols., Moscow, 1962, 1963.
S. P. Tolstow, Auf den Spuren der altchoresmischen Kultur, Berlin, 1953.
Turkish language studies. General and Old Turkish: C. Brockelmann, Osttürkische Grammatik der islamischen Literatursprachen Mittelasiens, Leiden, 1954.
S. Čağatay and S. Tezcan, “Köktürk tarihinin čok önemli bir belgesi: Soguṭča Bugut yazıtı,” TDAY-Belleten, 1975-76, pp. 245-52.
G. Clauson, An Etymological Dictionary of Pre-Thirteenth-Century Turkish, Oxford, 1972.
G. Doerfer, “Die özbekischen Lehnwörter in der Sprache der Araber von Buchara,” Central Asiatic Journal 12, 1969, pp. 296-308.
Idem, “Bemerkungen zur Methodik der türkischen Lautlehre,” Orientalistische Literaturzeitung 66/7-8, 1971, pp. 325-44.
Idem, “Das Chorasan-türkische,” TDAY-Belleten, 1977, pp. 127-204.
Idem, “Materialien zu türk. h-,” Ural-altaische Jahrbücher, N.S. 1, 1981, pp. 93-141; 2, 1982, pp. 138-68.
Idem, Lexik and Sprachgeographie des Chaladsch, Wiesbaden, 1987a.
Idem, “Maḥmūd al-Kāšγarī, Aṛγu, Chaladsch,” Ural-altaische Jahrbücher, N.S. 7, 1987b, pp. 105-14.
Idem, “Chaghatay,” in EIr. (forthcoming).
J. Eckmann, Chagatay Manual, Bloomington, 1966.
A. von Gabain, “Iranische Elemente im zentral- and ostasiatischen Volksglauben,” Studia Orientalia 47, 1977, pp. 57-70.
Idem, “Irano-Turkish Relations in the Late Sasanian Period,” in Camb. Hist. Iran III/1, 1983, pp. 613-24.
Gadāʾī, The Dīvān of Gadāʾī, ed. J. Eckmann, Bloomington, 1971.
G. Jarring, An Eastern Turki-English Dictionary, Lund, 1964.
Maḥmūd Kāšḡarī, Compendium of the Turkic Dialects, ed. R. Dankoff and J. Kelly, 3 vols., Harvard, 1982-85.
B. Khasanov, Yazyki narodov Kazakhstana i ikh vzaimodeĭstvie, Alma-Ata, 1936.
Moḥammad Mahdī Khan, Sanglax. A Persian Guide to the Turkish Language by Muhammad Mahdī Xān. Facsimile Text with an Introduction and Indices by Sir Gerard Clauson, London, 1960.
M. Mansuroglu, “Das Karachanidische,” in Philologiae Turcicae Fundamenta I, pp. 87-112.
K. H. Menges, The Turkic Language and Peoples, Wiesbaden, 1968.
Faḵr-al-Dīn Moḥammad Manṣūr (Faḵr-e Modabber Mobārakšāh), ed. E. Denison Ross, Ta’ríkh-i Fakhru’d Dín Mubáraksháh. Being the Introduction to the Book of Geneologies of Fakhru’d-Dín Mubáraksháh Marvar-rúḍí Completed in A.D. 1206, London, 1927.
Orkun Namık, Eski Türk yazıtları, Ankara, repr. 1987.
Philologiae Turcicae Fundamenta, 2 vols., Aquis Mattiacis [Wiesbaden], 1959, 1964.
W. Radlotf, Versuch eines Wörterbuches der Türk-Dialecte, 4 vols., repr. The Hague, 1960.
A. Saparov, Vzaimootnosheniya tyurkskikh yazykov khorezmskogo oazisa, Tashkent, 1988.
Karakalpak: N. A. Baskakov, Karakalpakskiĭ yazyk, 2 vols., Moscow, 1952 (with valuable notes on dialects).
Idem, Karakalpaksko-russkiĭ slovar’, Moscow, 1958 (best dictionary).
K. H. Menges, Qaraqalpaq Grammar I, New York, 1947.
D. S. Nasyrov et al., Bibliograficheskiĭ ukazatel’ po karakalpakskomu yazykoznaniyu (1925-1976 gg.), Nökis, 1978.
Kazakh: M. B. Balaqaev and T. R. Qordabaev, Grammatika kazakhskogo yazyka, 2 vols., Alma-Ata, 1967 (best grammar).
I. Bukin, Russko-kirgizskiĭ i kirgizsko-russkiĭ slovar’, Tashkent, 1883.
A. Ïsqaqov et al., Qazaq tiliniŋ qïsqaša ètimologiyalïq sözdigi, Alma-Ata, 1966.
V. Katarinskiĭ, Russko-kirgizskiĭ slovar’, Orenburg, 1890.
S. K. Kenesbaev, ed., Mestnye osobennosti v kazakhskom yazyke, Alma-Ata, 1973.
J. R. Krueger, Introduction to Kazakh, Bloomington, 1980.
Kh. Makhmudov and G. Musabaev, Kazakhsko-russkiĭ slovar’, Alma-Ata, 1954 (best dictionary).
G. G. Musabaev, Sovremennyĭ kazakhskiĭ yazyk I: Leksika, Alma-Ata, 1959.
Idem, ed., Qazaq tiliniŋ dialektologiyalïq sözdigi, Alma-Ata, 1969.
Ḡ. Qaliev and Š. Sarïbaev, Qazaq dialektologiyasy, Alma-Ata, 1967.
L. Z. Rustemov, Qazirgi qazaq tilindegi arab-pansy kirme sözderi, Alma-Ata, 1982.
Sh. Sh. Sarybaev, Bibliograficheskiĭ ukazatel’ literatury po kazakhskomu yazykoznoniyu I, Alma-Ata, 1965.
B. N. Shnitnikov, Kazakh-English Dictionary, London, The Hague and Paris, 1966.
Khorasani Turkish: A grammar and a dictionary by Hesche and Doerfer are in preparation; much material is found in Arazkuliev et al. and Berd’ev et al. (see Turkmen, below), and cf. Ivanow (below, under Khorasani Persian. For now see Doerfer, 1977.
Kipchak-Uzbek: See Oghuz-Uzbek, below.
Kirghiz: R. J. Hebert and N. Poppe, Kirghiz Manual, Bloomington, Ind., 1963.
G. Imart, Le Kirghiz turk d’Asie centrale soviétique, 2 vols., Aix-en-Provence, 1981.
Zh. Mukambaev, Dialektologicheskiĭ slovar’ kirgizskogo yazyka I, Frunze, 1976.
A. Tursunov et al., Ocherki grammatiki i leksiki kirgizskogo yazyka, Frunze, 1965.
K. K. Yudakhin, Kirgizsko-russkiĭ slovar’, Moscow, 1965 (excellent dictionary).
B. M. Yunusalev, Kirgizskaya leksikologiya I, Frunze, 1959.
Oghuz-Uzbek: There is no complete grammar or dictionary of this language, but Abdullaev, 1961, 1967 (with a map of the distribution of dialects), Šoabdurrahmonov, and Dobos (see Uzbek, below) may be consulted.
Tatar: A. M. Memetov, Istochniki formirovaniya leksiki krymsko-tatarskogo yazyka, Tashkent, 1988.
Turkmen: S. Arazkuliev et al., Kratkiĭ dialektologicheskiĭ slovar’ turkmenskogo yazyka, Ashkhabad, 1977.
P. Azimov and Kh. Baĭliev, Türkmen diliniŋ grammatikasï, Ashkhabad, 1963.
N. A. Baskakov et al., Turkmensko-russkiĭ slovar’, Moscow, 1968 (best dictionary).
N. A. Baskakov et al., Grammatika turkmenskogo yazyka, Ashkhabad, 1970.
R. Berdyev et al., Ocherk dialektov turkmenskogo yazyka, Ashkhabad, 1970.
B. Char’yarov and G. Sarev, Grammatika turkmenskogo yazyka, Ashkhabad, 1977.
O. Hanser, Turkmen Manual, Vienna, 1977.
Uighur (modern): A. Baskakov and V. M. Nasylov, Uĭgursko-russkiĭ slovar’, Moscow, 1939.
Sh. Kibirov and Yu. Tsunvazo, Uĭgursko-russkiĭ slovar’, Alma-Ata, 1961.
E. N. Nadzhip, Modern Uigur, Moscow, 1971.
Uzbek: F. A. Abdullaev, Khorezmskie govory uzbekskogo yazyka, Tashkent, 1961.
Idem, Fonetika khorezmskikh govorov, Tashkent, 1967.
A. K. Borovkov, ed., Uzbeksko-russkiĭ slovar’, Moscow, 1959.
Zh. Buronov et al., English Uzbek Dictionary, Tashkent, 1977.
É. Dobos, “An Oghuz Dialect of Uzbek Spoken in Urgench,” Acta Orientalia Hungarica 28, 1974, pp. 75-97.
E. I. Fazylov, Starouzbekskiĭ yazyk. Khorezmiĭskie pamyatniki XIV veka, 2 vols., Tashkent, 1966, 1971.
Idem and L. G. Chichulina, Russkie tyurkologi i uzbekskoe yazykoznanie, Tashkent, 1979.
A. von Gabain, Özbekische Grammatik, Leipzig and Vienna, 1945.
A. N. Kononov, Grammatika sovremennogo uzbekskogo literaturnogo yazyka, Moscow and Leningrad, 1960 (best grammar).
M. Mirzaev, “Bukhoro ŭzbek va tojik ševalarining uzaro munosabati tŭḡrisida,” in Materialy po uzbekskoĭ dialektologii II, ed. V. V. Reshetov, Tashkent, 1961.
V. P. Nalivkin and M. V. Nalivkina, Russko-sartovskiĭ sartovsko-russkiĭ slovar’, Kazan, 1884 (still useful).
A. H. Nauta, “Der Lautwandel von a > o and von a > ä in der özbekischen Schriftsprache,” Central Asiatic Journal 16, 1972, pp. 104-18.
A. Raun, Basic course in Uzbek, Bloomington, 1969.
A. M. Shcherbak, Grammatika starouzbekskogo yazyka, Moscow and Leningrad, 1962.
A. Sh. Shukurov and D. Bazarova, Uzbekskoe sovetskoe yazykoznanie. Bibliograficheskie ocherki (po 1982 g.), Tashkent, 1986.
F. Sjoberg, Uzbek Structural Grammar, Bloomington, Ind., 1963.
Š. Š. Šoabdurahmonov, Ǔzbek ḵalq ševalari luḡati, Tashkent, 1971.
N. Waterson, Uzbek-English Dictionary, Oxford and New York, 1980.
S. Wurm, Der özbekische Dialekt von Andidschan, Brno, etc., 1945.
Other languages: N. A. Baskakov and A. I. Inkizhekova-Grekul, Khakassko-russkiĭ slovar’, Moscow, 1953.
For additional bibliographies see, for Uzbek: Kenesbaev, pp. 41-57, Mirzaev, Qaliev and Sarïbaev, p. 81, Shukurov and Bazarova, pp. 18-19, 220; for Karakalpak: Nasyrov et al., pp. 27-28, 51-53; for Kazakh: Ïsqaqov, Musabaev, 1959, pp. 55-57, 107-25, Rüstemov, Sarybaev, pp. 147, 154, Qaliev and Sarïbaev, pp. 80-81 (Iranian loanwords); for Kirghiz: Yunusaliev, pp. 219-30; for Turkmen and Uzbek/Oghuz: Saparov.
Iranian languages (see also xiii, above): H. W. Bailey, “A Turkish-Khotanese Vocabulary,” BSOAS 11, 1943-46, pp. 290-96.
Ch. Kh. Bakaev, Govor kurdov Turkmenii, Moscow, 1962.
J. Benzing, Das chwaresmische Sprachmaterial einer Handschrift der “Muqaddimat al-Adab” von Zamaxšarī I: Text, Wiesbaden, 1968.
E. E. Bertel’s, editor-in-chief, Tadzhiksko-russkiĭ slovar’, Moscow, 1954 (the best dictionary).
G. Doerfer, Türkische Lehnwörter im Tadschikischen, Wiesbaden, 1967.
D. I. Edel’man, Yazgulamsko-russkiĭ slovar’, Moscow, 1971.
A. A. Freĭman, Khorezmiĭskiĭ yazyk. Materialy i issledovaniya 1, Moscow and Leningrad, 1951.
W. B. Henning, “Mitteliranisch,” 1958. Idem, A Fragment of a Khwarezmian Dictionary, ed. D. N. MacKenzie, London, 1971.
W. Ivanow, “Notes on the Ethnology of Khorasan,” Geographic Journal 67, London, 1926, pp. 143-58.
Idem, “Notes on Khorasani Kurdish,” Journal and Proceedings of the Asiatic Society of Bengal, N.S. 23, 1927, no. 1, pp. 167-235.
Idem, “Persian as Spoken in Birjand,” JASB, N.S. 24, 1928, pp. 235-351.
L. N. Kiseleva and V. I. Mikolaĭchik, Dari-russkiĭ slovar’, Moscow, 1978.
D. N. MacKenzie. The Khwarezmian Element in the Qunyat al-munya, London, 1990.
I. M. Oranskiy, Die neuiranischen Sprachen der Sowjetunion, 2 vols., The Hague and Paris, 1975.
A. Pagliaro and A. Bausani, Storia della letteratura persiana, Milano, 1960.
P. Pelliot, “Les influences iraniennes en Asie Centrale et en Extrême-Orient,” Revue d’histoire et de littérature religieuses N.S. 3/2, March-April 1912, pp. 97-119.
V. S. Rastorgueva, Tadzhiksko-russkiĭ dialektnyĭ slovar’, Ocherki po tadzhikskoĭ dialektologii 5, Moscow, 1963.
Idem, Opyt sravnitel’nogo izucheniya tadzhikskikh govorov, Moscow, 1964.
A. Z. Rozenfel’d, Tadzhiksko-russkiĭ dialektnyĭ slovar’ (Yugo-vostochnyĭ Tadzhikistan), Leningrad, 1982.
G. Sköld, Materialien zu den iranischen Pamirsprachen, Lund, 1936.
V. S. Sokolova, Ocherki po fonetike iranskikh yazykov, 2 vols., Moscow, 1953.
Z. V. Togan, Documents on Khorezmian Culture, pt. I: Muqaddimat al-Adab with the Transation [sic] in Khorezmian, Istanbul, 1951.
Arabic: W. Fischer, “Die Sprache der arabischen Sprachinsel in Uzbekistan,” Der Islam 36, 1961, pp. 232-63.
I. N. Vinnikov, “Slovar’ dialekta bukharskikh arabov,” Palestinskiĭ sbornik 10 (73), Moscow and Leningrad, 1962.
Originally Published: June 30, 1991
This article is available in print.
Vol. V, Fasc. 3, pp. 226-235
Gerhard Doerfer, “CENTRAL ASIA xiv. Turkish-Iranian Language Contacts,” Encyclopædia Iranica, V/3, pp. 226-235, available online at http://www.iranicaonline.org/articles/central-asia-xiv (accessed on 30 December 2012).
Turkic is a large family of about 40 languages stretching from Turkey all the way to China. Most of the languages are pretty close, and it’s often been said that they are all mutually intelligible, and that you can go from Turkey all the way to the Yakut region of Siberia and be understood the whole way.
This is certainly not the case, although there is something to it. That is because the languages, while generally not above 90% mutually intelligible which is the requirement to be dialects, do have varying degrees of intelligibility. That is, there is some intelligibility between most of the Turkic languages but generally below 90%.
The truth is that mutual intelligibility in Turkic is much less than proclaimed.
Azeri is spoken in Azerbaijan. Turkish and Azeri are often said to be completely mutually intelligible, but this is not true, though the situation is interesting. The two are not mutually intelligible. The far eastern dialects of Turkish are closer to Azeri than to Turkish. Turkish has an average of 69% intelligibility with Azeri calculated via three separate studies. After a few weeks of close contact, they can often communicate pretty well. Written intelligibility is much higher and Turks may have up to 95% intelligiblity of written Azeri.
Intelligibility is increasing now now due to increased contact. Nowadays due to exposure to Turkish TV, most Azeri speakers can speak Turkish well, and due to exposure to Azeri TV, Turks understand a lot more Azeri than they used to.
Kazakh and Kirghiz are also close, enough to be one language, with intelligibility over 90%. In addition, they have been growing closer recently. Kazakh is spoken in Kazakhstan, and Kirghiz is spoken in Kyrgyzstan.
Tatar and Bashkir are even closer than Kazakh and Kirghiz and they are best seen as a single language, with intelligibility of over 90%.
Uzbek and Uyghur are fairly close, but they are still probably only 65-70% intelligible. Uzbek is spoken in Uzbekistan, and Uighur is spoken Xinjiang Province, China.
Uzbek and Kazakh are not mutually intelligible, but there is an intelligible dialect between them.
Tofa and Tuvan are not mutually intelligible, but there are intelligible dialects linking them. Both are spoken in Russia in the same region as Altai below.
The truth is that Altai and Uzbek are not even intelligible within themselves.
Altai is spoken in the Altai region of Russia where China, Russia and Mongolia all come together. Altai is split into North Altai and South Altai, separate languages.
Uzbek is split into North Uzbek and South Uzbek, separate languages.
Azeri is split into North Azeri and South Azeri, although the two are mutually intelligible, there are large differences in phonology, morphology, syntax and loan words. Nevertheless, they are very mutually intelligible, with intelligibility at 98%. The split was probably done for political reasons, as North Azeri is the official language of Azerbaijan and South Azeri is a language spoken in Northwest Iran.
The Oghuz languages are said to be fully mutually intelligible, but that’s not really the case. The question of the intelligibility of Turkmen with Azeri and Turkish is controversial, as some sources say that they are mostly mutually intelligible. Intelligibility testing is warranted.
Turkish has uncertain intelligibility with Crimean Tatar. Crimean Tatar speakers say that Turks cannot understand their language (Dokuzlar 2010). However, Turkish speakers say that Turks and Crimean Tatar speakers can converse without too many problems. However, while mutual intelligibility is high, it is probably under 70%. Intelligibility testing is warranted. One problem is that Southern Crimean Tatar is a simply a dialect of Turkish, while Central and Northern Crimean Tatar are part of a separate language from Turkish.
Turkish has high, but not full, intelligiblity of Karaim. Turkish intelligibility of Karaim may be 65-70%. Intelligibility testing is warranted.
The intelligibility of Turkish with South Azeri may be quite high, on the order of 90% (however, some South Azeri speakers say that while they can understand North Azeri just fine, they have a hard time understanding Turkish, which calls the 90% figure into question), higher than between Turkish and North Azeri, which itself is ~70%. Intelligiblity between Turkish and South Azeri is the highest between Turkish and any other language.
The intelligibility of Turkish and Khorasani Turkic is probably around 40%.
Practically speaking, Turkish has low intelligibility with Kazakh (Kipchak Branch), Uyghur and Uzbek (Uyghuric branch) and Khakas (Siberian branch). Turkish-Kazakh intelligibility is surely less than 40%. There is also low intelligibility between Turkish and Bashkir, Nogay, Kyrghyz and Tatar (Kipchak Branch). Turkish has very low written intelligibility of Tatar (~5%) and Kazakh (0%).
Turkic has effectively 0% intelligibility with Yakut or Sakha.
The intelligibility of Turkish with the Central Asian Turkic languages like Uzbek, Kazakh, Kyrghyz and Turkmen is much exaggerated.
Speakers of these languages who went to study in Turkey said they had problems with the Turkish language. It’s true that Turkish TV is not much watched in the Central Asian Turkic nations, but the main reason for that is that Central Asian Turkic speakers can’t understand it. They can’t even understand the simplified Turkish used in these broadcasts. After the fall of the USSR, people from these new nations visited Turkey, but they had to bring interpreters with them to communicate.
In truth, the whole notion of the mutual intelligibility of all Turkish is a pan-Turkic conceit. Pan-Turkism is a noxious form of ultranationalism headquartered in Turkey. It says that all speakers of Turkic languages are part of a Greater Turkey and often uses ominous irredentist language implying that Turkey is going to conquer all the Turkic lands and take them back.
The Pan-Turkics have a snide attitude towards other Turkic speakers, insisting that they all speak dialects of Turkish and not separate languages. This snideness is resented by speakers of other Turkic tongues.
A number of Turkic languages are nothing more than dialects and not full languages.
Ukrainian Urum is a dialect of Crimean Tatar, and Georgian Urum is a dialect of Turkish. Ukrainian Urum is spoken in SE Ukraine, and Crimean Tatar is spoken on the Crimean Peninsula.
Salchuq is an Azeri dialect. It is spoken in Iran.
However, Qashqai, also spoken in Iran, often thought to be an Azeri dialect, is in fact a separate but closely related language with 75-80% intelligibility of South Azeri.
Gagauz has high intelligibility with Turkish. However, Bulgarians say that when Turks visit the Balkan Gaguaz communities in Bulgaria, the two groups have a hard time understanding each other. SIL says that not only Gagauz but also Balkan Gagauz Turkish are separate languages, but one wonders what criteria they are using to split them. The Gagauz are Christians living in Moldavia who strangely enough speak a Turkish language with many Christian Slavic loanwords. The Balkan Gagauz Turks live in Bulgaria, far west Turkey, Greece and Macedonia, but most of them live in Bulgaria.
Kumyk is said to be said to be intelligible with Azeri, which would make it a dialect of Azeri. However, this assertion is yet unproven, so for now, Kumyk should remain a separate language. Kalmyk is spoken in Dagestan.
Karakalpak is so close to Kazakh, with 98% intelligibility, that it is a dialect of Kazakh. Karakalpak is spoken in Western Uzbekistan.
Chulym and Shor are often thought to be dialects of a single language. Not only is this not true, but Shor itself is two separate languages – Mrass Shor and Kondoma Shor – and Chulym is also two separate languages – Lower Chulym and Chulym. Chulym and Shor are spoken north of the Altai Mountains in the Ob River Basin near the city of Novokuznetsk.
Further research regarding the intelligibility of these languages is indicated.
- Uygar Dokuzlar, Crimean Tatar speaker. April 2010. Personal communication.
If you think this website is valuable to you, please consider a contribution to support the continuation of the site.
Filed under Azerbaijani, Balkan Gagauz Turkish, Bashkir, Chulym, Crimean Tatar, Dialectology, Gagauz, Karakalpak, Kazakh, Khakas, Khorasani Turkic, Kipchak, Krymchak, Kumyk, Kyrgyz, Language Families, Linguistics, Nogay, North Altai, North Azerbaijani, North Uzbek, Oghuz, Qashqai, Salchuq, Shor, Siberian, South Altai, South Azerbaijani, South Uzbek, Tatar, Tofa, Turkic, Turkish, Turkmen, Tuvan, Urum, Uyghuric, Uyhghur, Uzbek