AFGHANISTAN v. Languages



v. Languages

It would be impossible to summarize here all the specialized research that has been carried out in linguistic and related ethnological fields. With few exceptions, dialectology and ethnology have proceeded independently; any attempt to provide a synthesis would reveal gaps at every stage and disparities that could not be resolved without the help of a vast undertaking such as the forthcoming Atlas linguistique de l’Afghanistane (ALA; ed. G. Redard, Berne). Only a cursory account can be given, a rough sketch delineating the present linguistic situation and the tribes that speak the various languages.

Situated at the intersection of three geographically and culturally different worlds—India with its monsoons, Central Asia with its steppes, and the Iranian plateau—Afghanistan has seen a succession of invaders and colonizers of all kinds. Its political history has been a constant battle for independence, its cultural history a struggle to maintain its own personality. States have appeared and disappeared, north, south, and straddling the Hindu Kush, but it has not been possible to confuse them with Central Asia, India or Persia. The Achaemenids (6th-4th cent. B.C.), Alexander and the Greeks (4th cent.), Aśoka and Buddhism (3rd cent.), Kanishka and the Kushans (1st cent. A.D.), the Sasanians (2nd-6th cents.), the Iranian Huns (4th-8th cents.) and the Hendūšāhīs of Kabul (1st-3rd/7th-9th cents.) demarcate pre-Islamic history. The coming of Islam (1st-3rd/7th-9th cents.) was the most important event in Afghan history. Islamic civilization flourished under the Ghaznavids (4th-6th/10th-12th cents.) and the Ghurids (6th-7th/12th-13th cents.), but the Mongol invasion in the 7th/13th century was a catastrophe from which Afghanistan never fully recovered. Nevertheless, the Timurid renaissance made Herat one of the great cities of the Islamic world in the 9th/15th century. In the 10th/16th century Bābor founded the dynasty of the Great Mughals. But the opening of the maritime route to the East Indies plunged the countries bordering the traditional silk route into economic and cultural stagnation. In the 12th/18th century Aḥmad Shah Dorrānī liberated Afghanistan from the influence of Persia and India and gave birth to modern Afghanistan.

Afghanistan’s ability to amalgamate rather than assimilate is such that it presents an extraordinary ethnic and linguistic medley. Ethnic diversity results from an agitated past, nomadism—which to this day is a way of life for about one million inhabitants (eight percent of the population)—and the geographic structure of the country. Around Afghanistan’s center of gravity, the Hindu Kush, are located great natural areas opening out on neighboring countries and lacking natural frontiers. The Āmū Daryā in the north, the desert of the west and the south, the mountain ranges in the east, are all passageways over which soldiers, missionaries, and merchants have traveled, while the central mountains are hospitable, having been populated since ancient times.

The linguistic situation (Table 7, Table 8). Best represented are the Iranian languages (see I. M. Oranskij, Iranskie jazyki, Moscow, 1963; tr. J. Balu, Les Langues iraniennes, Paris, 1977), followed by Turkish languages of recent import, and Indian languages which are either native (Nūrestānī and Dardic) or imported (New Indian). Most Afghans who are not native Persian speakers are more or less bilingual. The population of Afghanistan can be estimated at 12 million.

A. Official languages. Paṣ̌tō (1) promoted to the status of official language by royal decree in 1936, is the native tongue of 50 to 55 percent of Afghans; as a second language it is spoken by less than 10 percent of the population. The Paṣ̌tō-speaking areas are located in the east, the south, and the southwest of the country. Important colonies of nāqelīn (displaced populations) have settled in Bactria. Persian (2) is the language most spoken in Afghanistan. The native tongue of twenty five percent of the population, it is split into numerous dialects. Darī (q.v.) is a term long recommended by Afghan authorities to designate Afghan Persian in contrast to Iranian Persian; a written language common to all educated Afghanis, Darī must not be confused with Kābolī, the dialect of Kabul and surrounding areas that is more or less understood by eighty percent of the non-Persian speaking population and is fast becoming the nation’s koine.

Although Paṣ̌tō has enjoyed official favor, it is little propagated among Persian speakers. First, it is difficult to proceed from a less complex (more analytical) language such as Persian to a more complex (more synthetic) language such as Paṣ̌tō; and second, Paṣ̌tō has been poorly taught, despite the efforts of the Ministry of Public Instruction. The tendency is, rather, for Paṣ̌tūns to learn Persian in the course of their movements or during their military service, while the number of Persian-speakers who can express themselves in Paṣ̌tō remains stationary.

B. National languages. Since 1980 Balūčī (3), Ōzbēkī (23), Torkmanī (24), Pašaī (16), and Nūrestānī (in fact the Katī or Katə́ dialect, 12) have been officially promoted to the rank of national languages. During the autumn of 1981, each of these had a daily radio broadcast. At the Ministry of Public Education classroom texts for elementary school were being prepared; several such texts had already been printed in Ōzbēkī, Torkmanī, and Balūčī. Several issues of newspapers had also appeared in these languages.

C. Local languages. These are languages whose number of speakers, geographical extension, and communication role seem to be such that their existence is not threatened in the near future. They include the languages of the Pamir (4-9, 25) and Nūrestān (13-15, 17), languages firmly linked to ethnic groups of Indian origin who jealously maintain their identity (19-22), Brāhūī (29), and the idioms of various itinerant groups improperly joined together under the name Jat (32). Under this same heading can also be placed the argots and sabirs connected to professional or social groups, of which only two examples have been given (30, 31).

D. Residual languages. The sociolinguistic situation of languages in this group seems to condemn them to disappearance in the relatively near future; linguistic contacts have already shaken their original phonological and morphosyntactic structures. They include the language of the southeast, Parāčī (10) and Ōrmuṛī (11), and also Tīrō (18), Uiḡur (26), Moḡolī (27), and Arabic (28). Tīrō has already practically disappeared, and Ōrmuṛī, Moḡolī, and Uiḡur are rapidly becoming extinct. Parāčī still has more than 5,000 speakers, but the increasingly serious difficulties facing mountain agriculture will soon endanger it (Ch. M. Kieffer, Stud. Ir. 6, 1977, pp. 97-125). For the same reason the still-living idioms of the Pamir may perhaps soon regress, especially since bilingualism (clan language along with Persian) is already the general rule throughout the country.

The political upheavals that have taken place since 1973 have prompted the authorities to define a linguistic policy: promotion of the five national languages, development of classroom textbooks and radio programs, publication of newspapers, and, in the case of Pašaī and Nūrestānī (Katī), the transcription of languages which until now had never employed writing.

The complexity of languages and dialects, an image of the ethnic medley, testifies to the diverse cultural options that have, at various times in history, opposed or amalgamated different peoples, religions, and even states on the soil of present-day Afghanistan. For example, the semantic contrast between the words Skr. devá- “divine, celestial; god” (or daíva- “divine power, destiny”) and Avest. daēva- “demon (after the reform of Zarathushtra)” is reflected today in the remote descendants of these two languages: Aškūn dēi “god,” Waygalī “1. god, 2. oath” (< Indian: favorable connotation) on the one hand; Kābolī and Pašaī Lauṛowānī dēw “demon, ogre,” Paṣ̌tō lēwə́ and Pašaī Kačūrī Sālā dēu “wolf” (< Iranian: pejorative connotation) on the other. Another striking example can be found in the different names for the sun that testify to ancient beliefs: Sanglēčī ormózd, Eškāmšī rḗmuz represent the Iranian god Ahura Mazdā; Yidḡa-Munǰī mīˊra/o, like Persian mehr, Ōrmuṛī of Kaniguram meṣ̌r or Waṇecī mīr, evoke the god Mithra; Tirāhī sṹri, Gawar-bātī sūrīˊ, like Kalaša of Rumbūr sṹri, etc., go back to a form sṹriya- linked to the name of the sun god; Pašaī sur probably goes back to sūrī, another name for Kuntī, one of the wives of the sun god; Aškūn sōˊ Waygalī sōi, sȫ, Katī , etc., are based on a form saurīˊ- “wife of the sun.” All this shows the importance of linguistic paleontology for this as yet insufficiently known region.

Iranian languages (1-11). 1. Paṣ̌tō. (q.v.). An Iranian language of the east, perhaps a Saka dialect, Paṣ̌tō originated in the northern part of the eastern Iranian area and borrowed a great deal from Indo-Aryan. It is spoken in Afghanistan in the regions of Mašreqī, Kābolestān, Wardak, Lōgar, Paktīā (Paktyā), Ḡaznī, the south from Kaṭawāz to Farāh, where it is the majority language, the region of Herat, and Afghan Turkestan, i.e., Bactria, where in the course of the last century important implantations have continued to increase the Paṣ̌tō presence. Paṣ̌tō speakers call their language Paṣ̌tō; only Persian speakers call it Afḡānī or Awḡānī, language of the Afghans.

The great koine of the east and the south, Paṣ̌tō is distinct from the other great common language, Persian, by more archaic and complex morphological traits. For example, it has kept the masculine and feminine gender distinction in all nominal forms, together with an inflection of two and sometimes three cases.

Declension of the adjective xōẓ̌ “sweet”:

    masculine feminine
singular subject case xōẓ̌ xwaẓ̌a
  oblique case xwāẓ̌ə xwaẓ̌ē
plural subject case xwāẓ̌ə xwaẓ̌ē
  oblique case xwāẓ̌ō xwaẓ̌ō


Paṣ̌tō thus has six different forms where Persian has only one: xᵛoš (ḵᵛoš).

Another peculiarity of Paṣ̌tō is the ergative/ possessive/passive construction of transitive verbs in the past:

zə tā wahəm “I strike you,” tə mā wahē “you strike me,” mā tə wahəlē “I struck you,” tā zə wahələm “you struck me”

— first person pronoun, subject case: “I”

— second person pronoun sing., subject case: “you”

— first person pronoun, oblique case: “me”

— second person pronoun plur., oblique case: “you”

-əm — personal marker for the first person singular

-ē — personal marker for the second person singular

wah- — verbal stem for the present of wahəl “to strike”

wahəl- — verbal stem for the past of wahəl “to strike”

The literal sense of the personal pronouns and markers shows that wahəl-, the verbal stem for the past, has a passive meaning: mā tə wahəlē signifies literally “of me you were hit;” tā zə wahələm signifies literally “of you I was hit.” In the conjugation, the third person singular is not morphologically distinct from the third person plural; in contrast, the tense, mode, and aspect system is complex.

Because of its structure Paṣ̌tō unquestionably belongs to the Iranian language family, but its vocabulary reveals important borrowings from Indo-Aryan languages at various periods. It belongs to the eastern group of Iranian languages where the initial plosives b-, d-, and g- of the western group are generally replaced by the fricatives w-, δ (Paṣ̌tō l-) and γ- (see Table 9).

The Paṣ̌tō dialects are usually divided into “soft” and “hard,” a classification based on the differences in the treatment of ṣ̌, ẓ̌, ž (“soft”) and x, g, j (“hard”). In the table (Table 10 and Table 10 contd.) of Paṣ̌tō tribes and dialects, an intermediate group of Manǰanəy dialects has been introduced, characterized by the realizations (close to ich-laut), γ̌ (a strongly palatalized velar spirant), and ž:

  A (Maḡrebī) B (Manǰanəy) C (Mašreqī)  
“woman” ṣ̌əjə ҳəza xəja /ҳ/x
“beard” ẓ̌irə γ̌ira gira ž/γ̌/g
“mill” žrandə zranda ǰranda ž/z/ǰ
  ṣ̌-/ẓ̌-/ž ҳ-/γ̌-/z- x-/g-/ǰ-  


The actual situation is much more complex; the area of the B dialects is also that of many mixtures: and γ̌, and g, or x and γ̌, but never ṣ̌ and ẓ̌ (A dialects) or x and g (C dialects). The B dialects constitute a geographically intermediate zone, comprising a numerous and tribally mixed population, clearly separating Mašreqī or Jalālābatī from Maḡrebī or Kandahārī. They are not characterized by a generic pronunciation of the phonemes commonly used to differentiate dialect areas, but present a variety of divergences that are at times important and unquestionably original. While the B dialects present no trace of uniformity, they are characterized by specific phonetic traits and certain peculiarities in grammatical structure. This intermediate zone is probably the result of historical contacts and can not be considered a simple mixture of the other two groups.

2. Persian. The Persian of Afghanistan is generally designated Fārsī by the Tajiks and related ethnic groups, Pārsī by Paṣ̌tō-speakers, and Darī by the government. Dialect divergence is abundant, but not to the extent of preventing mutual comprehension, except for cases of occasional words and idioms that are easily clarified. Since contacts between Iranians and Afghans usually occur among speakers of relatively cultivated dialects, there are no particular problems in their mutual comprehension; the polished speech of both nations is fairly unified. Great difference is observable between local dialects and slang. Moreover there is a distinct cleavage between Iranian and Afghan Persian, on whatever speech level, in vocabulary. This applies not only to technical terms (Kābolī palås: Iran. gāz [ambor] “tongs”) but also to items of daily life (Kābolī maska: Iran. kare “butter;” bura: šakar “sugar”), familial terms (Kābolī dayi: Iran. māmā “midwife;” måmå: dāy “maternal uncle”), and the names of many objects imported from, or originating, abroad (Kābolī nektåi, from English “necktie”: Iran. kerāvāt, from French “cravate”; cf. Tajik galstūḵ, from Russian [and ultimately German “Halstuch”]).

Table 11 indicates the chief dialects. In general group A is more conservative than, e.g., the Persian of Tehran. This is noted in the phonology (maʿrūf and maǰhūl vowels), morphology and syntax (archaic uses of - and pronoun suffixes), and lexicon (survival of many Arabic words in active use). All have some Ōzbēkī elements. Dialects of the Kabul region may be distinguished generally from those of Afghan Turkestan by the presence of Indian elements in the former group and of Tajik influences in the latter.

Group B comprises those dialects which, in one way or another, share features with the dialects of Iranian Khorasan. E.g., one may draw comparisons between the speech of Herat and that of Mašhad. The group has some clear distinctions. On the central route, between Panǰāw and Ōba, when one leaves Garmāw going west and enters Aymāq territory, the local speech is readily distinguished from that of the Hazāras. As far as Garmāw, the present writer noted išpiš “louse,” muṛča “ant,” and čårmaḡz “walnut.” Beyond that point were recorded šubuš, muṛče/muṛčä, and ǰowz among the Aymāq, Fīrūzkōhī, and Taymanī. Similar forms occur up to Herat and Qaḷʿa-ye Now. The urban dialect of Farāh, while distinguishable from the speech of eastern Iran, is clearly affiliated with it; cf. a similar dialect among the Tajik Shiʿites of Qandahār.

Group C is chiefly represented by Sīstānī, a dialect straddling the Iranian-Afghan border. (It was designated by A. Ḡ. R. Farhādī by the name “Pahlavānī”.) Its center lies in Nīmrōz province, near the city of Zaranǰ at Kang (lit., “high ground” [emerging from the level of a swamp]). Balūčī influence on this dialect is marked.

Group D comprises the forms of Hazāragī spoken in Kāhmard, Bāmīān, Bēsūd, Nāwor, the west of Ḡaznī, Jāḡorī, Mālestān, Orūzgān, Gīzāw, Dāy-Kundī, Panǰāw, Yakāwlang, Šārestān, Laʿl-o-Sarǰangal. Their mutual contrasts are often only minor, and their typology remains to be fully defined. As a group, they are distinguished by such features as retroflex consonants and the presence of Mongol vocabulary items.

The vowel system of Afghan Persian differs from that current in Iranian Persian. The distinction between the maʿrūf vowels (the phonemes /i/ and /u/) and the maǰhūl vowels (/é/, /o/) is retained. Some minimal pairs appear in the following words, illustrating the vowel system (the phonetic realizations of the phonemes are in parentheses):

The vowel features which establish oppositions are qualitative (timbre) rather than quantitative (length) [Chart 1]. The allophones of /e/ (`ě, ǐ) and those of /o/ (`ǒ, ǔ) display the contrast between an urban or cultivated pronunciation (e.g., čelem “waterpipe,” bozorg “great”) and a rural or lax pronunciation (čǐlem or čǐlǐm, bǔzorg or bǔzǔrg). An epenthetic vowel (e, ə) dissociates consonant clusters (e.g., adel or adəl “justice”). In western Afghanistan, especially at Herat and near the frontier with Iran, words ending in -a (e.g., gofta,ḵåna) are closer to the Iranian forms (goftä/-gofte,ḵånä/ḵåne). The consonant system is nearly the same as in current Iranian Persian, with the exception of the contrast between the phonemes /q/ and /ḡ/, which is usually retained (qår “angry”: ḡår “cave;” båqi “remainder”: båḡī “pertaining to a garden”). Certain eastern dialects which have undergone contamination by Paṣ̌tō, and also some forms of Hazāragī, have developed a system of retroflex consonants. The stress accent is less prominent than in Iranian speech but otherwise presents no remarkable characteristics.

Various features of morphology are worth noting. (1) A complex system of demonstrative pronouns éni “this”: ónu “that;” ami ““this very”: amu “that very;” énami “this very (thing)”: ónamu “that very (thing),” and the plural forms: ényå : ónwå,amyå : amwå,énamyå : ónamwå. (2) Adverbs of place; of proximity: īǰa, énīǰa, amīǰa, énamīǰa “here, in this very place;” of distance: uǰa, ónuǰa, amuǰa, ónamuǰa “there, in that very place.” Adverbs of direction: isu, énisu, amisu, énamisu “toward here;” ósu, ónosu, amosu, ónamosu “toward there.” To these may be added the comparative suffixes . . . istar “more toward here;” . . . ustar “more toward there.” Adverbs of manner: étó(r), énetó, ametó, énametó “this way;” ótó(r), ónotó, amotó, onamotó “that way.” Adverbs of quantity: éqa, éneqa, ameqa, énameqa “this many;” óqa, ónoqa, amoqa, ónamoqa “that many.” (3) Numerals: du “two;” avda abda “17;” dosad/dosat “200;” yak-(k)am čel “39” (to avoid si-u-nó, homonymous with “complaisant cuckold”). (4) Prepositions: baḵč-e (-t) “for (you);” kat-e (-t) or kati (-t) “with (you).” (5) Postposition fatara “since”: (az) kay fatara “since when, how long?” (6) The enclitic morpheme -é/-e (Persian -ī, the yā-ye waḥdat) has a connotation of restriction, of concrete particularization (somewhat in the manner of an indefinite article): ketåbe méḵåna “he is reading a book”: ketåb mé ḵåḵåna “he reads.” (7) The plural morphemes, which are stressed, are -å and (rarely) -ån: mardå “men;” åqåyån-o ḵånomå “ladies and gentlemen.” (8) Personal pronouns: The third person plural forms are (*īnhā) and (*ānhā). The first person singular is ma and, for more familiar usage, mti (*man). (9) The personal pronoun suffixes are (1st-3rd pers. sing.): -em(a), -et(a), -eš(a); the suffix-a (*-ra) marks the sentence’s definite object. (10) Verbal endings -om (*-am, 1st sing.), -a (*-ad, 3rd sing.), -én (*-ēd, 2nd plur.), -an (*-and, 3rd plur.). (11) The preterite and the perfect are contrasted; e.g., for the verb raftan “go”:

ra̩ftom rafté̩m (*rafta̩-am)
ra̩fti rafti̩  
raft rafta̩ (*rafta̩-ast)
ra̩ftem raftem (*rafta-ēm)
ra̩ftén rafté̩n  
ra̩ftan rafta̩n  


(12) Dubitative constructions are formed with ḵå(t): ḵåt borom or ḵåt raftom “I may perhaps go;” rafta ḵåt budan “they have perhaps (already) gone.” (13) The durative prefix is or mó- in some rural dialects (*-): mékonom “I am doing;” mékadom “I was doing.” (14) The progressive periphrasis takes the form: da hål-e kår kadan astom “I am (in the act of) working;” cf. Iranian Persian dāram kār mīkonam. (15) In familiar usage, the suffix -ak is often attached to the verbal ending (e.g., raftomak “I went”).

At first glance the syntax of Afghan Persian does not seem to differ considerably from that of Iranian Persian, at least in its main lines. But there are many contrasts in detail. The enclitic morpheme - (Kābolī -ra/-a) is used more often in current Afghan speech. Afghan and Iranian usages contrast especially in that, to mark attribution, Afghan Persian uses -, while Iranian Persian forms phrases with the preposition be. Special uses of pronominal suffixes are noted: goft-em(-a) “he told me” (= ma-rā goft), goft-et(-a) “he told you” (= to-rā goft); mé-zani-š-a (*mē-zanī-eš-rā) “you are striking him;” mé-zanéš-ā (*mé-zana-eš-rā) “he is striking him.” Cf. the enclitic in constructions of the type: zan-e ḵód-eš-a sar-eš étébår nés “his own wife does not have faith in him” (lit., “for his own wife [zan-e ḵod rā] regarding him faith there is not”); gap-a na-rasidi “you have not understood” (lit., “you have not arrived at the statement” [*gap-rā]). A frequent type of complex sentence in Afghan Persian, with a subordinate clause of time, follows this pattern: padar-em ke åmad, ma maktab raftom “when my father arrived, I went to school” (Kābolī; cf. Iranian Persian: čūn padar-am āmad, man be madrase raftam).

3. Balūčī (q.v.). Though Balūčī does not have the numerical and cultural prestige of Paṣ̌tō or Persian, it is important enough for Radio Kabul to grant it a daily broadcast. Its ancestor is neither Parthian nor Middle Persian but a lost language sharing some features with both. In the genealogical table (Table 8) it has been put with Persian in the Western group because it is strongly marked by borrowings made from Persian since the time of Middle Iranian. The Balūčī claim the name Raḵšānī for their dialect. It is spoken in the area of the lower course of the Helmand, in Bagat and in Mīrābād (above, in Banāder, Safar, etc., and below in Ḵānnešīn and Dēšū, only Paṣ̌tō is spoken), in Kānī-Ḡar, Zīārat-e Bībī, Zīārat-e Šāhemsalī (Šāh-Esmāʿīl) and in the area of Kōh-e Malekdokān concurrently with Brāhūī. Then from Tāḡaz and Ḵᵛāǰa ʿAlī Soflā on, only Balūčī is spoken. This remains true along the course of the Helmand, in Palālak, Landəy, Dahmarda, Rōdbār, Qaḷʿa-ye Pādšāh, Qaḷʿa-ye Mādar-e Pādšāh, Čārborǰak, Qaḷʿa-ye Afżal, Bandar-e Kamālḵān, and Mīrābād; then, going towards the north, in Qaḷʿa-ye Fatḥ, Sabzgazī, Ḵᵛābgāh, Kang-e Dīnmohd, Zaranǰ, Zīārat-e Amīrān Ṣāheb, Kang, Deh-e Dōst Moḥammad, Kurki, Čaḵānsūr, and towards the northeast, in Kadō, Ḵās and Lōḵī.

4-9. Iranian languages of the Pamir. This appellation is used for dialects as different as Šuḡnī, Rōšānī, Eškāšmī, Sanglēčī, Munǰī (qq. v.; cf. Geiger and Kuhn, Grundr. Ir. Phil., 1/2, pp. 288-344). Though not an entirely satisfactory term (cf. the stricter classification of G. Redard, Current Trends in Linguistics VI, pp. 103-6), it is convenient from a geographical point of view: It groups the dialects which dominate the Pamir mountains and “border on Turkish languages on the one side and Indian dialects spoken by the peoples of the Hindu Kush on the other” (Geiger and Kuhn, I/2, p. 290). These dialects form part of the important group known as Northeastern Iranian languages; to them may be added Ossetic (spoken today in the Soviet Caucasus), Yaḡnōbī (the only residue of Sogdian, spoken in a transversal valley of the Zarafšān river southeast of Samarqand), the other Pamir dialects of the “Šuḡnī-Rōšānī group” (Baǰūī, Bartangī, Orōšōrī, Sariqōlī, Ḵūfī; see V. S. Sokolova, Ocherki po fonetike iranskikh yazykov. 2. Osetinskiĭ, yagnobskiĭ i pamirskie yazyki, Moscow, 1953; to these may be added Yazḡulāmī, a closely related dialect, and Wanǰī, today extinct; see G. Morgenstierne, Etymological Vocabulary of the Shughni Group, Wiesbaden, 1974, p. 5), Yidḡa (a variety of Munǰī spoken in Chitral), and Paṣ̌tō.

Northeastern Iranian dialects present sui generis characteristics, some of which can be listed here. With respect to historical phonetics, the initial voiced stops b-, d-, g- have generally changed to the corresponding fricatives v- (w-), δ- (then l-), γ- (then ž-). Thus Paṣ̌tō has w-, l- (< δ-), γ-; Eškāšmī and Sanglēčī have v-, γ-, but d-; Šuḡnī and Rōšānī have v-, δ-/d-, ž-/g-: Wāḵī has v-, δ-, γ- and Yidḡa and Munǰī have v-, γ-, but Munǰī has d- while Yidḡa has l- (see Table 9). Northeastern Iranian has the affricates c (ts) and j (dz) which are lacking in the dialects of Western Iranian. Thus to Avestan čaθwārō “four”, Persian čahār, Balūčī čār, correspond Paṣ̌tō calōr, Eškāšmī and Sanglēčī cəfúr, Šuḡnī cav/fṓr Rošānī cavūİœr, Wāḵī cībīˊr, but Munǰī čfūr. Northeastern Iranian is also characterized by certain grammatical features, such as the fact that the noun phrase develops towards the left (in the order: /determiner/-/determined/), while in the Western languages it develops towards the right (in the order: /determined/determiner/): Darī qālīn-e kōtá-ye má(n) “the rug of my room” ≠ Paṣ̌tō zmā də kōṭē γāləy idem, Sanglēčī mič tāt xān “our father’s house,” Šuḡnī mo dād čīd (andir) “(in) my father’s house,” Wāḵī ẓ̌e yaš mad “my horse’s back,” Munǰī wazīr lúγdo/a “the minister’s daughter.” The dialects of the Southeast, Parāčī and Ōrmuṛī, present the same trait.

From the historical point of view, little can be added to what has been written by G. Morgenstierne (Indo-Iranian Frontier Languages II. Iranian Pamir Languages, Oslo, 1938). In general, however, the following remarks can be made: (1) There is no linguistic or archeological basis for hypotheses concerning the origin of the populations and the date of their arrival in Šuḡnān, Sanglēč, Wāḵān (Waḵān), and Munǰān. (2) There are no indications concerning possible human settlements in this area during the historical period. (Such research would probably have to follow up that undertaken at Šortūgay by H. P. Francfort and M. -H. Pottier, “Sondage préliminaire sur l’établissement proto-historique harappéen et post-harappéen de Shortugaī,” Arts Asiatiques 34, 1978, pp. 29-79). (3) The Pamir dialects seem to have developed on the spot from Middle Iranian languages of which nothing is known. But philological evidence and linguistic structure would seem to exclude that any of them developed from Bactrian. (4) Present place-names have not yet furnished any indication of populations, perhaps of non-Iranian origin, that might have settled in this area before the arrival of the present inhabitants. (5) To resolve such problems research needs to be developed in three different directions: Much more linguistic material, including all place-names, needs to be collected, compared with existing materials, and analyzed linguistically and historically. A systematic archeological survey of the entire area needs to be conducted. Parallel research must be undertaken in such fields as cultural geography, ethnography, ethnology, economics, technology, sociology, and ethno-botany.

The future of the Pamir languages is clearer than their past. No doubt their destiny depends on complex economic, cultural, and political conditions, which will induce the speakers either to keep or to abandon their clan language, but on the whole the die is cast. These languages have no prestige, since they are spoken only by ethnic minority groups that are dependent militarily and economically on others. They are not religious languages; nor are they common languages, since no one learns or teaches them outside the family, the village, or the valley. All the speakers of the Pamir languages also know Persian; it is not without reason that they have long been called “Tajiks of the Pamir” or “Tajiks of the Mountains” (e.g., A. von Schultz, Die Pamirtadschik, Giessen, 1914, and idem, Landeskundliche Forschungen im Pamir, Hamburg, 1916; cf. Ch. M. Kieffer, “Einführung in die Wakhi-Sprache und Glossar,” Grosser Pamir, ed. R. Senarclens de Grancy and R. Kostka, Graz, 1978, p. 350). Their bilingualism will inexorably lead to the death of clan languages. The Pamir languages have been preserved in a remote mountain area that was largely self sufficient. But during the last decades the political and economic situation has changed rapidly and profoundly. The improvement of roads has considerably reduced traveling times. Mass media, particularly Radio Kabul, and civil servants, teachers, and soldiers have carried both Darī and Kābolī, along with a new citified concept of the world and life, into the Pamir homes. School teachers and other civil servants, when not foreign to the area, are trained in Kabul, or at least in Kondūz or Fayżābād. For the most part they exhibit the complexes of the semi-educated along with ignorance, shortsighted greed, and a profound contempt for all local culture. As a consequence of generalized military service, the native population is becoming increasingly mobile, and the exodus of the young is increasing, due to the attraction of well paid jobs in the city. One of the immediate consequences of this dislocation of institutions, customs, and traditional values is the abandoning of clan language, which will cause those who leave their native land to lose their identity in two or three generations. While the economic situation of the low and fertile land is improving, all the mountain areas of Afghanistan are undergoing a process of pauperization that will inevitably and in a short time bring about the rupture of traditional socio-economic balances and upset the status of clan languages. In some areas, such as the Waḵān for example, opium abuse accelerates the disintegration of these societies.

10-11. Southeastern Iranian languages. Ōrmuṛī and Parāčī (qq.v.), both spoken south of the Hindu Kush watershed, can be considered the last representatives of a linguistic group that enjoyed an important expansion before being submerged by Paṣ̌tō and Persian. Today Ōrmuṛī is spoken by fewer than a hundred Ōrmuṛ; in a few fortified qaḷʿa farms in the vicinity of Barakī Barak, in the province of Lōgar, and outside Afghanistan by some one thousand Bərkī or Brakī living among Paṣ̌tō speaking Wazīr and Masʿūd in Kaniguram in Pakistan, a village of southern Waziristan. Parāčī is spoken by ca. 5000 Parāčī in three valleys on the southern slope of the Hindu Kush: Šotol, between Sālang and Panǰšīr, and Ḡočūlān and Pačaḡān, both in the area of Neǰrao (Neǰrāb).

Since Ōrmuṛī and Parāčī are related languages, their belonging to the Eastern group of Iranian languages, or the Northwestern as Oranskiĭ and Payne have concluded, must be considered an essential problem. In Les langues du monde (Paris, 1952, p. 34) they are classified less than satisfactorily as dialects of the Pamir area. But G. Morgenstierne had indicated as early as 1926 (Report on a Linguistic Mission to Afghanistan, Oslo, 1926, pp. 14-37) that they form a distinct linguistic group whose essential traits are the following: With respect to phonetics, retention of initial voiced stops g-, d-, and b- as in Western Iranian languages (see Table 9). Development of v > γ/(w): Av. vafra- “snow,” Pers. barf, Paṣ̌tō wāˊwra, Orm. yōˊš/ṣ̌, Par. γarp. Development of-d > -γ: Orm. drȧy “long,” cf. Sanskrit dīrghá-; Par. maγas “fly,” cf. Pers. magas. Loss of -t- and -d-: Av. mātar- “mother,” Pers. mādar, Paṣ̌tō mōr, Orm. māˊwa, Par. . Development of -p- and -b- > -w-: Orm. tṓwa “sun,” cf. Av. tap- “to be warm;” Par. xṓwān “shepherd,” cf. Paṣ̌tō špūn. With respect to the lexicon, the correspondences are just as striking: Orm. g(i)ri “mountain,” Par. gir, ger “stone,” cf. Av. gairi-, Paṣ̌tō γar “mountain;” Orm. gap “stone,” Par. gapār, “hearth” (cf. R. L. Turner, A Comparative Dictionary of the Indo-Aryan Languages, London, 1966, no. 4023), *gabba-; Orm. undraw- “to sew,” Par. andarf-,idem; Orm. of Kaniguram “iron,” Par. ru, idem. With respect to morpho-syntactic structure, there are important divergences, probably because these two languages had separated before the breakdown of the ancient inflection system of Iranian. There are nonetheless convergence points, as for example the causative formation with -aw- in Ōrmuṛī and -ew- in Parāčī.

Whatever may be the linguistic problems raised on the one hand by the relationship between Ōrmuṛī and Parāčī, and on the other by ancient relationships between the Northeastern and Southeastern Iranian language groups, it is possible to acknowledge an ancient group of probably Southeastern Iranian languages located south of the Hindu Kush watershed, and represented today only residually by Ōrmuṛī and Parāčī. In ancient times these languages were spoken in much more extensive areas, as is shown by geographical, historical, toponymic, cultural, archeological, and epigraphical considerations (Ch. M. Kieffer, “La fin proche des langues iraniennes résiduelles du sud-est, ōrmuṛī et parāčī, en Afghanistan,” Langage et sociéte′ 10, December 1979, pp. 43-47). The region where Proto-Ōrmuṛī was spoken is perhaps delineated by the Barakestān of Bābor (Bābur-nāma in English, tr. A. S. Beveridge, 2nd ed., London, 1969, p. 220), the place-names ending in -grām, and by the trilingual inscription in stone at Dašt-e Nāwor (G. Fussman, “Documents épigraphiques kouchans,” Bulletin de l’Ēcole française d’Extrême-Orient 61, 1974, p. 34: the inscription DN III could have been made in Proto-Ōrmuṛī). This area would extend beyond Lōgar, to the east of Ḡaznī up to the place called Barakestān, and to the west up to the northwestern tip of Dašt-e Nāwor, while Kaniguram would represent a relatively recent migration. The Parāčī sounding place-names Estūfālō (the last hamlet in the valley of Šotol, situated at an altitude of 2360m) and Estālef (a village situated in Kōhdāman, some 50 km north of Kabul) can both be interpreted to mean “cowparsnip (Heracleum species of the Umbelliferae, Par. estūf) valley (Par. ālō);” they would extend the Parāčī speaking area towards Kabul considerably (Ch. M. Kieffer, “La fin proche,” pp. 45-47). If these hypotheses were to be confirmed, the two residual languages of the southeast would occupy two opposite triangular zones meeting at their summits in the area of Kabul. Parāčī in the north would occupy all of Kōhdāman and present-day Kōhestān up to the linguistic area of the Dardic and Nūrestānī languages. Ōrmuṛī, south of the Kabul river, would extend in the direction of present-day Hazāraǰāt in the west and the Solaymān mountains in the east. It was probably in these locations that the two conquering languages, Paṣ̌tō and Persian, attacked them and finally drove them back to their present territory. Other Southeastern languages may have existed, but they have disappeared without leaving any known trace.

The destiny of the Southeastern Iranian languages is even more irremediably fixed than that of the Pamir languages. Changes in the economic and socio-cultural situation have already upset their status to the point where their very existence is questionable. The Ōrmuṛ have forgotten their geographic origin and abandoned the religious traditions of the Rōšanīya of Bāyazīd Anṣārī (see EI2 I, pp. 1121-24), which distinguished them from neighboring ethnic groups and gave their clan language the prerogatives of a secret language. Beaten back by the imperatives of economic and technical development, they will soon have no choice but to join the ranks of the Paṣ̌tūn or the Tajiks who own the land, control the bazaars, and hold the key administrative positions (see Ch. M. Kieffer, Grammaire de l’ōrmuṛī, forthcoming).

Since 990/1582, as a result of Islamization, the Parāčī have experienced profound changes in their religious beliefs and ethics (see Ṣifat-nāma-yi Darvīš Muḥammad Ḫān-i Ġāzī. Cronica di una crociata musulmana contro i Kafiri di Laġmān nell’anno 1582, ed. G. Scarcia, Rome, 1965); for a long time these affected their language without really endangering it. However the isolation protecting them from the common languages gradually decreased as the political unification undertaken by Amir ʿAbd-al-Raḥmān (1302-19/1885-1901) progressed. More recently, construction of roads and compulsory military service have dealt a fatal blow. Economic development accomplished the rest; not only did it create new needs, upset the traditional network of market places, and break the balance of prices for agricultural and industrial products, it also permanently destroyed Parāčī self-sufficiency by drawing irreplaceable manual labor to the factories and the towns.

Indo-Aryan languages (12-22). Not surprisingly, the eastern edge of the Iranian plateau is marked by ancient and modern contacts with the Indian sub-continent. Languages of the Indo-Aryan family spoken in Afghanistan can be divided into two main groups: Nūrestānī and Dardic, having their home in present Nūrestān, and neo-Indian, spoken by ethnic minorities of Indian origin more recently implanted in the area. The Nūrestānī and Dardic languages have been studied in particular by G. Grierson (Linguistic Survey of India X. Specimens of the Languages of the Eranian Family, 1921; repr. Delhi, 1968), G. Morgenstierne (Norsk tidsskrift for sprogvidenskap 2, 1929; 7, 1934; 12, 1940; 13, 1945; 15, 1949; 16, 1952; 17, 1954; 18, 1958; Acta orientalia 8, 1930; 12, 1934; 18, 1939; Report on a Linguistic Mission to Afghanistan, Oslo, 1926; Report on a Linguistic Mission to North-West India, Oslo, 1932; Indo-Iranian Frontier Languages III. Pashai, 1-3, Oslo, 1967, 1944, 1956; etc.), G. Buddruss (Kanyawali, Proben eines Maiya-Dialektes aus Tangir [Hindukusch], Munich, 1959; Beiträge zur Kenntnis der Pašai-Dialekte, AKM 33/2, Wiesbaden, 1959; Die Sprache von Woṭapūr und Kaṭārqalā, Bonn, 1960; Die Sprache von Sau in Ostafghanistan, Beiträge zur Kenntnis des Dardischen Phalūra, Munich, 1967), and R. F. Strand (“Notes on the Nūristānī and Dardic Languages,” JAOS 93, 1973, pp. 297-305; Strand retains much unedited material). G. Fussman gives the best overview of these languages and a complete survey of the problems they present in his Atlas linguistique des parlers dardes et kafirs (Paris, 1972). The term “Dardic and Kafir languages” is inconvenient (ibid., pp. 11-14), since kāfer (“infidel”) is not appropriate to designate Moslems; it should be replaced by “Nūrestānī” (see R. F. Strand, “Notes,” p. 297). The only “infidels” in this region of Asia are the Kalaš, who do not live in Nūrestān, but in Chitral, and who speak a Dardic language, Kalaša (see G. Morgenstierne, Indo-Iranian Frontier Languages IV. The Kalasha Language, Oslo, 1973).

12-15. The Nūrestānī languages. These four languages, to which it might be appropriate to add Trēgāmī (but not Zamyākī, which is only a form of Waygalī), are all spoken in Nūrestān, inside the borders of present-day Afghanistan. They form a linguistic group that is sometimes considered intermediary between the Iranian and Indo-Aryan groups. Although their vocabulary is largely Indo-Aryan and their phonetic evolution has followed a course similar to that of the Dardic languages, they also present a certain number of non-Indian characteristics, such as the total loss of the aspirate trait in their phonological system and the survival of the ancient Indo-European distinction between palatalized velar and labio-velar stops. Also found are residual archaic traits, such as the survival of the dental -s- after u, and specifically Nūrestānī lexical items, attested neither in Indo-Aryan nor in Iranian (Morgenstierne, Report, 1926, pp. 50-69 and Fussman, Atlas, pp. 12-13). One is therefore justified in postulating the existence of either a third branch of Indo-Iranian, or an ancient, pre-Vedic off-shoot of Indo-Aryan; the Nūrestānī languages of today would constitute a small residual group, rebuilt with the help of Indo-Aryan materials of various periods, but still presenting a few traces of the ancient group (Fussman, Atlas, p.13; Morgenstierne, Irano-Dardica, Wiesbaden, 1973, pp. 327-43).

16-18. The Dardic languages. These three isolated languages of the mountains are clearly Indo-Aryan, but unlike the Indo-Aryan languages of the plains, they have preserved archaic traits in their vocabulary and phonological structure; the latter is characterized by numerous consonant groups and the preservation of three types of voiceless spirants: the dental s, the retroflex (ṣ̌) and the palatal š (see Morgenstierne, Indo-Iranian Frontier Languages III/1-3; idem, Report, 1926, pp. 81-93; Fussman, Atlas, p. 12). The most important of these languages with respect to the number of speakers is Pašaī, which was recently given the status of a national language (see G. Buddruss, Beiträge, 1959). Gawar-bātī appears in three varieties: Šumaṣṭī, spoken in an isolated point (Šumaṣṭ) in Pašaī country (G. Morgenstierne, NTS 13, 1945, pp. 239-81), Gřˊangalī, signaled by W. Lentz in 1937 (Deutsche im Hindukusch, p. 273) and studied by G. Buddruss in 1970 (G. Fussman, Atlas, p. 25) and by A. L. Gryunberg (Indiĭskaya i iranskaya filologiya, Moscow, 1971, pp. 329), and Ningalāmī, which now seems to be extinct. The Dardic group also stretches into Pakistan (Chitral, Swat, etc.), where it comprises many dialects.

19-22. The Neo-Indian languages. These testify to the establishment of various Indian ethnic groups at least as early as the Ghaznavid period. Sikh and Hindu temples can be found in Kabul and most of the other cities of eastern Afghanistan. Sikh merchants speak Panǰābī (19), while other Indian minority groups speak Sindhī (20) and Inku (Lahndā, 22). Some Guǰur cattle breeders in Konar speak Goǰrī (Guǰurī, 21; see ALA). Dialects of the Sirāikī (Sindkhī) type, about which relatively little is known, are used by the Vaṛŋgāwālā, the Pikrāǰ, the Šādībāz, and the Jalālī; these peoples along with other groups are given the totally improper label Jat (see 32, below).

Turco-Mongolian languages (23-27). Ōzbēkī (23) and Torkmanī (24) are especially used in an area of northern Afghanistan between the Āmū Daryā in the north, Persian speaking Badaḵšān in the east, Kōh-e Bābā and the Paropamisus in the south, and the basin drained by the Morḡāb in the west. Ancient Bactria, the cradle of important Indo-Iranian establishments, of Greco-Bactrian kingdoms, and of the Kushan Empire that gave fame to the Bactrian language, has always been exposed to invasions from Central Asia. The Ōzbēk invasion that has gradually submerged it began in the 10th/16th century, while the Turkmans came especially in the 19th and beginning of the 20th century, fleeing the Russian advance in Central Asia. Today the Afghans call Bactria Turkestan, although the autochthonous bulk of the population is Tajik, i.e. Persian speaking. The most numerous Turkish speakers are the Ōzbēks, whose language belongs to the eastern group, though it has been so deeply influenced by Iranian languages that, e.g., vowel harmony has entirely disappeared (A. von Gabain, Özbekische Grammatik, Leipzig and Vienna, 1945). The Turkmans, chiefly settled in the northwest along the Āmū Daryā and the Soviet border, speak a dialect of the Oḡuz group (see L. Bazin, “Le Turkmène,” in Philologiae Turcicae Fundamenta I, Mainz, 1959, pp. 308-17; G. Jarring, On the Distribution of Turk Tribes in Afghanistan, Lund and Leipzig, 1939). Many Turkish speakers, especially Ōzbēks, have migrated, mainly in the direction of large cities like Kabul (see D. Balland, “L’Immigration des ethnies turques à Kabul,” in E. Grötzbach, ed., Aktuelle Probleme der Regionalentwicklung und Stadtgeographie Afghanista n , Meisenheim am Glan, 1976, pp. 210-24). They retain their clan language, despite the inevitable bilingualism (Persian), as long as they live in groups or maintain close contacts (often economic in nature) with parent groups. The women given in marriage to Tajik and especially Paṣ̌tō men only exceptionally transmit their Turkish tongue to their children.

25. Qirḡizī. It is spoken in the Pamir by a Qirḡiz population belonging essentially to two tribes: Teyit and Kesek (see R. Dor, Contributions à l’étude des Kirghiz du Pamir afghan, Paris, 1975). The Qirḡiz occupation of parts of the Pamir, intermittent as early as the 18th century, changed to permanent settlements towards the end of the 19th century. Their dialect is part of the Qipčaq group (G. Hazai, EI2 V, s.v. Ḳipčaḳ); it is marked by contacts with Tajiki and Kābolī Persian (ALA, questionnaire normal, 6.b, collected by R. Dor).

26. Uiḡur. It is still spoken in Kabul by a small minority of “Uighurized” Afghans, repatriated from Chinese Xinjiang in the early 1960s (Balland, “L’Immigration,” p. 212). Since they live in contact with the Ōzbēks of the capital, their dialect is marked by Kābolī Persian and Ōzbēkī (Ch. M. Kieffer, Le parler uighur de Caboul, forthcoming).

27. Moḡolī. In Afghanistan it is a residual language spoken in the area of Herat (Kūndūr, Kārēz-e Mollā, Bedawī, Deh-e Šayḵ, Dū-Rūdī, Samanābād, Naw [Nāb near Ōbē]; M. Weiers, Die Sprache der Moghol der Provinz Herat in Afghanistan, Opladen, 1972). Moḡolī materials have been gathered elsewhere, but these are clearly of lesser importance and have often remained unpublished (ibid., pp. 11-13; ALA questionnaire réduit 200 et questionnaire normal 233; use H. F. Schurmann, The Mongols of Afghanistan, The Hague. 1962, only with caution (cf. K. Ferdinand, Acta Orientalia 28, 1964, pp. 175-203]). The written sources of Afghanistan Moḡolī have been published by W. Heissig (Schriftliche Quellen in Moġolī I. Texte in Faksimile, Opladen, 1974) and M. Weiers (Schriftliche Quellen in Moġolī II. Bearbeitung der Texte, III. Poesie der Mogholen, Opladen, 1975-77). It is not clear why a few Moḡolī speakers have kept their clan language. Sh. Hattori’s hypothesis that it probably served as a secret language is worth retaining (Studies in General and Oriental Linguistics, Presented to Shiro Hattori, ed. R. Jakobson and Sh. Kawamoto, Tokyo, 1970, p. xiii). Moḡolī dialects seem to have been more widely spoken, as is suggested by the existence of many Persian speaking Moḡol groups who have abandoned their clan language but have not forgotten their origins; moreover, the Mongol substratum in Hazāragī Persian is tenacious (G. K. Dulling, The Hazaragi Dialect of Afghan Persian, London, 1973). These Moḡolī dialects were probably quite diversified from the point of view of dialectology and chronology. But today Moḡolī is almost extinct; profoundly disorganized in its vocabulary and its syntax because of its contacts with Persian, it survives mainly as a language of folklore and memories.

Other languages. 28. Arabic. The Arabic-speaking Arabs (q.v.) of Afghanistan live in western Bactria to the west of Mazār-e Šarīf in four villages of 50 to 100 houses each: Ḥasanābād, 14 km east of Šeberḡān; Solṭān Areḡ, 12 km east of Āqča; Yaḵdān, 4 km southwest of Dawlatābād-e Balḵ; and Ḵōšḥālābād, 5 km south-southwest of the same village. All the men speak Arabic, Persian, Ōzbēkī, Torkmanī, and sometimes even Paṣ̌tō. Persian is the cultural and religious language of Bactria and has affected the Arabic of the region, as have the Ōzbēkī and Torkmanī of the Turkic-speaking agriculturalists. The women’s speech is less contaminated, as is usual in the case of domestic languages (see Ch. M. Kieffer and R. Kieffer-Vonmoos, “Notes de dialectologie arabe, I. A propos du parler des femmes de Ḥasanābād et de Solṭān Areġ en Bactriane,” Mélanges Maxime Rodinson, forthcoming). From the linguistic point of view, the qəltū Arabic of Bactria, like that of Central Asia described by W. Fischer (“Die Sprache der arabischen Sprachinsel in Uzbekistan,” Der Islam 36, 1961, pp. 232-63) is closest to the dialects of Mesopotamia (Handbuch der arabischen Dialekte, ed. W. Fischer and O. Jastrow, Wiesbaden, 1980, pp. 140-54), but it presents generic peculiarities in its phonetics and morphosyntax, some of which do not belong to the structure of Arabic and indicate important contact phenomena. It is extraordinary that these Arabs of the Qorayš tradition have kept their clan language. In the past they must have had precise reasons for maintaining their Arabic identity, but they are incapable of defining them today. Established in Bactria towards the end of the 8th/14th century, perhaps by Amir Tīmūr himself, they were probably guardians of the Timurid order (Ch. M. Kieffer, “L’arabe et les Arabophones de Bactriane, I. Situation ethnique et linguistique,” Die Welt des Islams 20, 1981, pp. 178-96). These Arabs are unrelated to the many Persian-speaking “Arab” groups.

29. Brāhūī. Dravidian is represented in Afghanistan by the Brāhūī dialect of the Brāhūīkān groups of Afghan Sīstān, mainly along the Helmand. They live practically in symbiosis with the Balūč and claim to belong to the same ethnic group; they have borrowed many lexical items from them. Brāhūī dialectology materials have been collected in Afghanistan by the staff of the ALA (questionnaire normal 222 and questionnaire réduit 221 and 224) and by Ch. M. Kieffer (Plantes et drogues d’Afghanistan, forthcoming).

30. Zargarī and other argots. Argots are used in many bazaars. Forms of Zargarī have been collected by Ch. M. Kieffer in Tāšqorḡān and Fayżābād (forthcoming); in both cases encoding processes are used (e.g., with -zar- and -šel-). Other argots (of butchers, rug merchants, etc.) have never been collected.

31. Lāzemī. Lāzemī is a sabir spoken mainly (but not only) in the Kabul area; merchants, artisans, servants, and generally all those who are in contact with foreigners use it to speak with them. The word itself is formed from the Arabic-Persian expression lāzem (ast) “(it is) necessary,” used loosely to mean “I want, I wish, I need, etc.” Thus barāy šomā lāzem ast “it is necessary that you (buy, bring, take, etc.)” or “You need” is one of the most frequent phrases. Its suffix -ī is the same as that found on the names of most other languages (Fārsī, Parāčī, etc.). A characteristic trait of Lāzemī is its extremely simplified verb conjugation; the second person of the imperative is the only form used: borū “go!”; man borū “I go,” tū borū “you go (sing.),” . . . šomā borū “you go (plur.),” etc.; dīrūz man borū “yesterday I went,” fardā man borū “tomorrow I will go,” etc. Lāzemī materials have been collected by Ch. M. Kieffer in Kabul and elsewhere (forthcoming). Concerning the social status of Lāzemī, see Studia Iranica 6, 1977, p. 112.

32. Various languages are used by itinerant linguistic groups improperly known as Jat (materials have been collected by A. Rao and will be published in collaboration with Ch. M. Kieffer). Some are of Indian type (see above, 19-22), while others are structurally Iranian, though not necessarily of the same status. These include the following: Ādūrgarī is a kind of secret language used by the Persian-speaking Šayḵ-Moḥammadī while engaging in commercial or artisan activities. The name (ādūr- “pedding” + the suffix -gar-ī of nouns of action or trade, cf. zar-gar-ī “trade of the goldsmith”) evokes precisely their major activity as itinerant peddlers (see A. Olesen, “The Sheikh Mohammadi, A Marginalized Trading Community in East Afghanistan,” in The Other Nomads . . , ed. A. Rao, forthcoming). Qazūlagī or Magadī (in the Herat region), also called Ḡorbatī, is the language of the Ḡorbat; A. Rao considers it their native tongue, even though it has features in common with Ādūrgarī, which is a secret language (Les Gorbat d’Afghanistan. Aspects économiques d’un groupe itinérant “Jat”, Paris, 1982). Magatibai is the language of the Jōgī, who move about in the provinces of Balḵ, Jōzǰān, Fāryāb, and Kondūz. Part of its vocabulary is shared by Magadī of the Ḡorbat (see A. Rao, “Note préliminaire sur les Jat d’Afghanistan,” Studia Iranica 8, 1979, p. 144). Similar languages have been mentioned by W. Ivanow as “jargons” of gypsies and “mendicant darwishes”. (Journal of the Asiatic Society of Bengal, N.S., 10, 1914, pp. 439-55; 16, 1920, pp. 281-91 ; 18, 1922, pp. 375-83; 23, 1927, pp. 243-45).


See also J. Elfinbein, “Report on a Linguistic Mission to Helmand and Nimruz,” Afghan Studies 2, 1979, pp. 39-44.

G. Fussman, “Documents épigraphiques kouchans,” Bulletin de l’Ēcole française d’Extrême-Orient 61, 1974, pp. 1-66, plates I-XXXIII.

A. L. Gryunberg, Yazyk kati, Moscow, 1980.

H. Humbach, Baktrische Sprachdenkmäler, 2 vols., Wiesbaden, 1966-67.

Karta narodov peredneĭ Azii, Moscow, 1960.

Ch. M. Kieffer, “Ētudes Parāči,” Studia Iranica 6, 1977, pp. 97-125, 249-81; 7, 1978, pp. 81-107, 269-77; 8, 1979, pp. 67-106, 245-67; 9, 1980, pp. 99-119, 233-49; 10, 1981, pp. 283-306.

W. Lentz, Pamir-Dialekte. I. Materialien zur Kenntnis der Schugni-Gruppe, Göttingen, 1933.

G. Morgenstierne, Indo-Iranian Frontier Languages: I. Parachi and Ormuri, Oslo, 1929.

II. Iranian Pamir Languages (Yidgha-Munji, Sanglechi-Ishkashmi and Wakhi), Oslo, 1938.

III. The Pashai Language, 1. Grammar, 2. Texts and Translations, 3. Vocabulary, Oslo, 1967, 1944, 1956. IV.

The Kalasha Language, Oslo, 1973 (This last contains a bibliography of Morgenstierne’s numerous works on Afghan linguistics).

Idem, Irano-Dardica, Wiesbaden, 1974.

Idem, “The Linguistic Stratification of Afghanistan,” Afghan Studies 2, 1979, pp. 23-33.

E. Orywal, Die Balūč in Afghanisch-Sīstān. Wirtschaft und soziopolitische Organisation in Nīmrūz, SW-Afghanistan, Berlin, 1982.

E. Yarshater, “Iran and Afghanistan,” Current Trends in Linguistics. VI. Linguistics in South West Asia and North Africa, ed. T. A. Sebeok, The Hague, 1971, pp. 669-89.

On Afghan Persian see D. C. Phillott, Higher Persian Grammar for the Use of the Calcutta University, Showing Differences between Afghan and Modern Persian, Calcutta, 1919.

G. A. Grierson, ed., Linguistic Survey of India X. Specimens of Languages of the Eranian Family, Delhi, 1921 (repr. 1969), pp. 452-53 (Dēhwārī), 527-30 (Badaḵšī).

D. L. R. Lorimer, The Phonology of the Bakhtiari, Badakhshani, and Madaglashti Dialects of Modern Persian, London, 1922. Morgenstierne, Report, 1926, pp. 7-9.

Idem, “Persian Texts from Afghanistan,” Acta Orientalia 6, 1928, pp. 309-28.

L. Bonelli, “Appunti fonetici sul volgare persiano di Kabul,” Annali dell’Istituto Universitario Orientale di Napoli 1, 1928-29, pp. 5-14; 2, 1930, pp. 24-26; 4, 1931, pp. 20-33; 8, 1936, pp. 43-53.

L. Bogdanov, “Stray Notes on Kābulī Persian,” Journal and Proceedings of the Asiatic Society of Bengal 26, 1930, pp. 1-125.

L. N. Dorofeeva, Obraztsy sovremennoĭ afganskoĭ pressy na kabuli, Moscow, 1952.

Idem, Putevoditel’ po Afganistanu s kratkim slovarem na kabuli, Moscow, 1952.

Idem, Opyt leksiko-grammaticheskoĭ kharakteristiki kabuli, Moscow, 1955.

A. Ḡ. R. Farhādī, Le persan parlé en Afghanistan. Grammaire du kâboli, accompagnée d’un recueil de quatrains populaires de la region de Kâbol, Paris, 1955.

J. C. Wilson, Jr., An Introduction to Colloquial Kabul Persian, Army Language School, Presidio of Monterey, Calif., 1958.

Dorofeeva, O yazyke farsi Afganistana, Moscow, 1959.

Idem, Yazyk farsi-kabuli, Moscow, 1960.

L. N. Kiseleva and V. I. Mikolaĭchik, Dari-russkiĭ slovar’, 21 000 slov, Moscow, 1978.

Afḡānī-nevīs, Loḡāt-e ʿāmīāna-ye fārsī-ye Afḡānestān, Kabul, 1340 Š./1961.

Wilson, One Hundred Afghan Persian Proverbs, mimeo., Kabul, 1961.

Idem, An Afghan Persian-English Dictionary, ms., Kabul. I. M. Oranskiĭ, Iranskie yazyki, Moscow, 1963, pp. 119-22, ed.

A. V. Rossi, Le lingue iraniche, Naples, 1973, pp. 133-36.

T. N. Pakhalina, “On the System of Vowel Phonemes in Kabuli-Persian (Some Results of Experimental Investigation),” Papers Presented by the USSR Delegation, XXVIth International Congress of Orientalists [New Delhi, 1964], Moscow, 1964.

Idem, “K kharakterstike Kabul’skogo prostorechiya,” Indiĭskaya i Iranskaya Filologiya, Moscow, 1964.

E. H. Glassman, Conversational Dari. An Introductory Course in Dari (= Farsi = Persian) as Spoken in Afghanistan (revised edition of “Conversational Kabuli Dari,” with the assistance of M. Taher Porjosh), Kabul (The Language and Orientation Committee, International Afghan Mission, P.O. Box 625), 1970-72.

T. Nawata, A Primer of Spoken Kabul Persian, Hiroshima, 1972.

G. K. Dulling, The Hazaragi Dialect of Afghan Persian, London, 1973.

D. M. Dōst, Da Afḡānestān žəbē aw tokəmuna, Kabul, 1354 Š./1975, pp. 326-50.

Farhādī, The Spoken Dari of Afghanistan. A Grammar of Kāboli Dari (Persian), Compared to the Literary Language, mimeo., Kabul, 1975.

M. A. Fekrat, Loḡāt-e zabān-e goftārī-e Herāt, Kabul, 1355 Š./1976.

Ch. M. Kieffer, Afghanica III. Textes kāboli (in preparation).


(Ch. M. Kieffer)

Originally Published: December 15, 1983

Last Updated: July 22, 2011

This article is available in print.
Vol. I, Fasc. 5, pp. 501-516

Cite this entry:

Ch. M. Kieffer, “AFGHANISTAN v. Languages,” Encyclopaedia Iranica, Online Edition, 1982, available at