IRAN vii, continued

vii(9). Arabic

During Sasanian times, several large Arab tribes moved into the Fertile Crescent. Some tribes were moved to the Persian hinterland by Shapur II (r. C.E. 309-79) to the area of Bam, others to Dälaki in Fars and Kerman. Beginning in the early seventh century, administrators and troops with families as well as tribes moved or were moved to cities, towns, and countryside in all regions of the former empire. Settlement occurred in various waves, developing in a pattern of immigration and subsequent assimilation to local varieties of Iranian and later also Turkic, differing from region to region. Most extensive was the Arab settlement in eastern Iran and Greater Khorasan (including northwestern Afghanistan, and Central Asia, including Marv and Bukhara). At its height, the total population of Arab immigrants is estimated to have totalled about a quarter of a million.

Most of the original settlers have lost their native language. The Arab origin may still be remembered (sometimes erroneously) and be reflected in tribal names, often of sub-tribes of larger confederations, and in toponyms. In order to indicate the erstwhile extent of the Arab settlement, the detailed descriptions by E. Daniel, P. Oberling, and B. Hourcade on present-day tribes, Arab-speaking or not, may be summarized as follows; cf. also ʿAbbāsi (1999) on Arabs in Khorasan; Sehampur (1999) on the Arabs of the Khamsa in Fars; and the comprehensive overview by ʿA.-Ḥ. Zarrinkub on the Arab conquest in the Cambridge History of Iran (1975):

(1) Khuzestan: Arab tribes in the province and along the southern littoral of the Persian Gulf.

(2) Fars hinterland: (a) the Il-e ʿArab of the Khamsa tribal confederacy; (b) two branches of the ʿArab Jabbāra and ʿArab Sheybāni, speaking a mixture of Arabic, Persian, and Lori; (c) further, the formerly independent Arab tribes, the Bahāʾ-al-dini and the Shiri, now absorbed by the Qashqā’i tribal confederacy and the Arab tribe of the Khamsa tribal confederacy, respectively.

(3) Kermān province (some Arabs of Kerman province are said to have still spoken Arabic in the 1930s): (a) Sirjān region, the ʿAṭāδallāhi; (b) Bardir (Māshiz), the ʿArabkhāni Sorkhi and three Badu’i tribes; Pāriz region, the ʿArab-e Ḥāji Ḥosayni and a fourth Badu’i group.

(4) Sistan: the Mir ʿArab and Sayyed tribes.

(5) Makrān and Persian Baluchestan: the Pishi and Mand on the Pakistan border, the Rend (doubtful) and the Kalmati.

(6) Lorestan: several of the Lori and South Kurdish Lakki tribes.

(7) Kurdistan: several of the main clans in Kermān-shāhān and Māhidasht.

(8) Azerbaijan: two tribes of the Shahsavan, and a clan of the Qaraqalpaq, who are now Turkicized.

(9) Mazandaran: a small group in the eastern province, and sayyeds in Gorgan and Fenderesk.

(10) Tehran province: semi-nomadic tribes in the Tehran region, the ʿAraba, now diminished by the sprawl of the capital, in the dehestān Jalilābād south of Varāmin, still transhumant; around Tehran, a heterogeneous amalgam of tribal fractions, deported there at one time or another, the Bāqeri, possibly a splinter of the Khamsa; the Mishmāst from Khuzestan; and the Koti from Shiraz, now dispersed around Varāmin.

(11) Central Iran, several groups: (a) some clans around Qom and Kāshān; (b) the Ardestāni tribe between Kāshān and Nāʾin; (c) further south, the ʿArab tribe of the Haft Leng of the Bakhtiāri tribal confederacy.

(12) Khorasan, with a large Arab population: (a) Shahrud area (still called ʿAjam o ʿAjam, probably the Bastami tribe); (b) the Arabs in the dehestān of Taḥt-Jolga, west of Nishapur; (c) the Il-e ʿArab tribe in the dehestān of Kandakli, west of Sarakhs, where Arabic was still spoken in two villages in the 1950s (Farhang IX, pp. 309, 350); (e) the Arabs around Torshiz (Kāshmar) and Torbat-e Jām, mostly ʿArab Mishmāst; (f) the Arabs around Torbat-e Ḥeydari, between Torshiz and Torbat-e Jām, where also Arabic was still spoken in the 1950s (Farhang IX, p. 86); (g) the Arabs around Tun (Ferdows); (h) the Arabs in the dehestān of Zir Kuh, between Qāyen and the Afghan border, where also Arabic was still spoken in the 1950s (Farhang, pp. 227, 345, 418); (i) the Arabs of the dehestān ʿArabkhāna, southwest of Birjand, where also Arabic was still spoken in the 1950s; (j) the Arabs in two villages in the neighboring dehestāns of Nahārjāt and Nehbandān.

In present-day Iran Arabic speakers are found in the areas adjacent to Arabic-speaking countries, mostly Khuzestan and the Persian Gulf region (between 200,000 to 500,000), in Fars, possibly Kerman, and in three pockets in Khorasan.

Linguistically, the dialects of Khuzestan are part of the continuum of Iraqi dialects. Their linguistic setting and features were outlined by B. Ingham (1973, 1976), who also identified Persian and Turkish loans in northeastern Arabia (Ingham, 2005) and an overview of the languages of the Persian Gulf (Ingham 1980).

The remnant Central Asian Arabic has been subject to linguistic research since the 1930s, in recent years foremost by Jastrow as the leading scholar of Arabic and Neo-Aramaic dialects in the Fertile Crescent (see discussion and references in Jastrow, 2005).

However, no linguistic data on the Arabic of Khorasan had been known until the field work by U. Seeger (2000) and S.-O. Dahlgren (2002-2003, 2005). They thereby could verify the informations in the Farhang of the 1950s on the Arabic-speaking villages in Khorasan (Sarakhs, Torshiz and Torbat-e Jām, Tun in Zir Kuh, and ʿArabkhāna of Birjand; see above). Dialectologically, they identified three small areas with villages fully or partially speaking Arabic: (1) the dehestān ʿArabkhāna southwest of Birjand; (2) Zirkuh, 100 km northeast of Birjand at the Afghan border; (3) Sarakhs at the far northeastern border with Turkmenistan, apparently emigrants from ʿArab-khāna. The latter two communities also identify themselves as ʿArab-e Khazāʿi. Dahlgren also found Arabic-speakers in the Māhidasht near Persepolis.

The linguistic material in these pioneering articles on Khorasan Arabic is necessarily limited, and leaves many questions open. The grammatical descriptions include brief sections on phonology, morphology, and some notes on syntax, and a total of five short sample texts, of which Dahlgren’s are provided with interlinear translation. It is hoped that all recorded materials will be published in due time.

In terms of Arabic dialectology, they have retained features by which they can be related to Mesopotamian (Iraqi) dialects prior to the so-called qeltu/gelet split, and thus to the immigrations during the ninth and tenth centuries (Dahlgren, 2002-03, 2005), based on the criteria that were established for Central Asian Arabic by O. Jastrow (1998).


Phonology. A feature inherited from the Mesopotamian source dialects is the shift of the interdental fricatives θ δ δ̣ > s z z (as in Central Asian Arabic; cf. Ratcliffe, 2005). The major later change is the loss of phryngealization, except há and ʿ (ʿayn). Q and k merge; dentals show strong palatalization before i e a: a-ctǟ “the book.”

Morphology. Much of the inherited morphology and morphophonology has been retained. Thus, masculine and feminine gender are distinguished in the nominal and verbal system in both second and third persons singular and plural. However, particularly the noun phrase and sentence syntax differ radically from the common Arabic typology.

Nominal system. Case distinction (nominative, genitive, accusative) is lost. Plural formations by suffix or ablaut (broken plural) are retained, and extended to loans: mīz > əmyǟz “tables,” kešvär > kešǟver besides kešver-ǟt “countries.” The definite article is al-/a-, which assimilates to dentals: ar-rūziyy “the child,” and is fully assimilated to the initial consonant in Zir Kuh: ab-bājir-a “the cow” (< baqara), which shows the feminine marker -a(t). The indefinite article färd (< fard “individual”) is innovated: färd emräyye “a, one woman” (cf. Persian yek). The personal pronouns and suffixes and the deictic pronouns are as follows: Chart 1.

Verbal system. The ablaut system defining the aspect forms and nominal forms, and derived stem forms (derived stem classes II, III, V, VII, VIII) have been retained to a considerable degree. Person and gender are marked by affixation: Chart 2.

Progressive forms found are formed, as in most Arabic dialects (already classical), with *kāna + present: kun ni-rāʿī-hinne “we (ni-) used to let them (3rd plur. fem.) graze.” Typical for these dialects, as for the Central Asian dialects, is the use of the present participle to express a perfect tense/aspect form corresponding to the Persian perfect (raft-a am; for Central Asian Arabic, cf. Windfuhr, 2005): masc. lāgt-unnä-h “he has picked him (3rd sing. masc. -h) up” < *lāgiṭ-in-hu; fem. āxiz-ṯ-inn-ah “I took her (-ah) [as wife]” < *āxiz-at-in-hu. As in other Arabic dialects, the formant -inn- shows the derivation of these suffixed forms from the generalized nunation -in. In turn, the inherited Arabic perfective (“preterit”) corresponds to the Persian preterit (raft-am). The general negation is ma-: mä-ya-ʿref-īn “they don’t know.”

The expression of the copula continues a common innovation. In 3rd person forms it is marked by the personal suffixes: dē xošmez-aw “this is delicious,” aksar-hum fi kermān-hum “most of them are in Kerman.” It can be marked for all persons by Persian existential hast- as follows (with halāk “tired”): Chart 3.

Syntax. Noun phrase. Definiteness is marked by al-/a- prefixed to the dependent noun and copied onto the adjective: ʿurubiyy-et el-kešver-ät el-uxra “the Arabic of the other countries,” dīc ä-mäntaγ-et el-uxra “that other region.” Indefiniteness is marked by -in; it is attached to the head noun: kətǟb-in /kutub-in zēn “a good book/good books,” mamǟlik-in qadīmiyy-ä “the former countries,” with feminine -a of the adjective. As evidenced by masal-in mā hū “it’s no problem” (cf. Persian mas’ala’i nist), this development of the Arabic nunation corresponds to the Persian indefinite -i, attached to both singular and plural: ketāb-i, ketāb-hāʾi “a, some certain book/books.”

Word order. Word order has shifted from VSO to Persian SOV: aḥne fi-j-jidīm māldār kun-ne “formerly we were herders” (< fi al-qadīm); kul-na ʿurubiyyä nə-hanžem “we all speak Arabic” (< H-N-J-M “to speak”).

Complex sentences. Unlike Central Asian Arabic, Khorasan Arabic does not appear to have copied the universal Persian conjunctive ki; at least it does not occur in texts. Instead, definite-specific relative clauses are introduced by the definite article: manteγ-et el-eḥna hast-i-na, ʿArabxāne, ḥudūd fi säd kilomētr-ēt Birjand “The area that we are in, Arabxāne, is approximately 100 km from Birjand.” Note the absence of a copula, and the calque on Persian dar sad kilometr-í-ye Birjand. Persian conjunctional phrases are regularly calqued: baʿd-a el-tarayyag-t “after I had breakfast”; cf. Persian baʿd az in ke.

Lexicon. The lexicon and phraseology freely borrow from Persian, including compound verbs, and usually shift to the phonology and articulation of Persian: färq yə-sännä “they (3rd plur. fem.) are different,” lit. “they make a difference.”


Bibliography of Arabic:

Bo Utas, “Semitic in Iranian,” in E´va Ágnes Csató, Bo Isaaksson, and Carina Jahani, eds., Linguistic Convergence and Areal Diffusion, London and New York, 2005, pp. 65-77.

Arabs in Iran. Elton L. Daniel, “ʿArab iii. Arab Settlement in Iran,” in EIr. II, pp. 210-14.

Pierre Oberling, “ʿArab iv.1. Arab Tribes of Iran,” in EIr. II, pp. 215-19.

ʿAli Ašraf Ṣādeqi, “Arabic i. Arabic Elements in Persian,” in EIr. II, pp. 229-31.

Aḥmad Tafażżoli, “Arabic ii. Iranian Loanwords in Arabic,” in EIr. II, pp. 231-23.

ʿAbd al-Ḥosain Zarrīnkūb, “The Arab Conquest of Iran and its Aftermath,” in Camb. Hist. Iran IV, 1975, pp. 1-56.

Tehran area. Bernard Hourcade, “The Arabs of Tehran Province,” in EIr. II, pp. 219-20a.

Hušang Sehāmpur, Tāriḵča-ye ilāt wa ʿašāyer-e ʿArab-e Ḵamsa-ye Fārs, Shiraz, 1999.

Khorasan. H. ʿAbbāsi, “Negāh-i ba pišina-ye tāriḵi-e aʿrāb-e janub-e Ḵorāsān,” in M. ʿAzizi, ed., Ḵorāsān-šenāsi, Tehran, 1999.

Sven-Olof Dahlgren, “Arabs in Central and Eastern Iran,” Orientalia Suecana 148, 2002-03, pp. 89-94.

Idem, “On the Arabic of Arabkhane in Eastern Iran,” in Csató, Isaaksson, and Jahani, eds., Linguistic Convergence and Areal Diffusion, 2005, pp. 161-71.

Ulrich Seeger, “Zwei Texte im Dialekt der Araber von Chorasan,” in Werner Arnold and Hartmut Bobzin eds., “Sprich doch mit deinen Knechten Aramä-isch, wir verstehen es!” 60 Beiträge zur Semitistik. Festschrift für Otto Jastrow zom 60. Geburtstag, Wiesbaden, 2002, pp. 629-46.

Fars. Bernard Hourcade, “The Arabs of Tehran Province,” in EIr. II, pp. 219-20a.

Khuzestan. Bruce Ingham, “Urban and Rural Arabic in Khuzistan,” BSOAS 36, 1973, pp. 533-53.

Idem, “Regional and Social Factors in the Dialect Geography of Southern Iraq and Khuzistan,” BSOAS 39, 1976, pp. 62-82.

Idem, “Languages of the Persian Gulf,” in Alvin J. Cottrell, ed., The Persian Gulf States: A General Survey, Baltimore, 1980, pp. 314-33.

Idem, “Persian and Turkish Loans in the Arabic Dialects of North Eastern Arabia,” in Csató, Isaaksson, and Jahani, eds., Linguistic Convergence and Areal Diffusion, 2005, pp. 173-79.

Central Asia. Otto Jastrow, “Die Position des Uzbek-Arabischen,” in Holger Preisler and Heidi Stein, eds., Annäherung and das Fremde: XXVI Deutscher Orientalistentag vom 25. bis 29.9. 1995 in Leipzig, Stuttgart, 1998, pp. 173-84.

Idem, “Uzbekistan Arabic: A Language Created by Semitic-Iranian-Turkic Linguistic Conversion,” in Csató, Isaaksson, and Jahani, eds., Linguistic Convergence and Areal Diffusion, 2005, pp. 133-39.

Robert R. Ratcliffe, “Bukhara Arabic: A Metatypized Dialect of Arabic in Central Asia,” in Csató, Isaaksson, and Jahani, eds., Linguistic Convergence and Areal Diffusion, 2005, pp. 141-59.

Gernot Windfuhr, “Central Asian Arabic: The Irano-Arabic Dynamics of a New Perfect,” in Csató, Isaaksson, and Jahani, eds., Linguistic Convergence and Areal Diffusion, 2005, pp. 111-23.

(Gernot Windfuhr)

Originally Published: December 15, 2006

Last Updated: December 15, 2006

This article is available in print.
Vol. XIII, Fasc. 4, pp. 401-404