ix. The Median Dialects of Kashan  

This sub-entry is divided into two sections: 

 (1) Rural Rāji dialects

 (2) Urban Jewish dialect.   


 (1) Rural Rāji Dialects 

Although the city of Kashan itself is now entirely Persophonic (see part 2), many of the settlements in its proximity have preserved their native Central dialects. The Central dialects of Kashan, which we call here the Kashan dialects as a whole, are often called Rā(ye)ji or Dei (Dehi) by their speakers (Yarshater, 1985; idem, 1989). The Rāji-speaking villages and townships associated historically with Kashan extend as far as Delijān in the west, Meyma in the southwest, and Naṭanz in the south, although these mentioned localities now form administrative districts independent from Kashan within the province of Isfahan. In fact, the current administrative divisions serve no good frame of reference for delineating dialects of what has been called the Kashani subgroup of the Central dialects (see also isfahan xxi). One of the aims of this article is to improve isoglottic definition of the Kashani group. 

In the past few decades, rural Kashan has rapidly been shifting to Persian. Most villages have already been partly or entirely persianized, and practically all Rāji speakers are bilingual (see Borjian, 2009). A distribution of the Rāji-speaking places is known from a survey conducted in the 1970s (Purriāḥi et al.) for individual rural districts of Isfahan Province. The data pertaining to Kashan, calculated and summarized in Table 1, shows that 38 of the total 96, or 40 percent, of the settlements had native Rāji speakers. In Kavirāt, a garmsir sub-district centered at Abuzaydābād, all nine settlements spoke Rāji, whereas among the eight settlements of Ārān Sub-district only two, namely Ārān and Bidgol, preserved, if only partly, their dialects. On the sardsir or mountainous south and west of Kashan District, Qohrud and Niāsar show substantial language shift, while Jowšaqān and Meyma have been conservative. The survey excludes the Naṭanz district south of Kashan, with several Rājispeaking villages around the Karkas peak, such as Hanjan, Yārand, Abyāna, Bidhand, Fariz(h)and, So(h), Ṭarq, Ṭār, Keša, as well as the garmsir sub-district of Bād(rud) (q.v.; see also isfahan xx).

In spite of the random distribution of the Rāji- and Persian-speaking settlements, and the profound influence of Persian on the Rāji dialects, there is no language continuum between the two, in the way that, for instance, Caspian blends into Persian as one moves south from central Alborz onto the plateau. The dialect discontinuity becomes evident when we compare adjoining villages, such as Rāji-speaking Ḵonb and Persian-speaking Latḥor (south of Kashan city) and Rāji-speaking Vandāda and Persian-speaking Ḵosravābād (south of Meyma), respectively in the north and the south of Kashan linguistic province. 

The following comparative data (from Purriāḥi et al., IV, 245-46; idem, II, pp. 189-90) demonstrates the point at hand. For “snow,” “boy,” “dog” we have Latḥori and Ḵosravābādi barf, pesar, sag, Ḵonbi vāfr, pirā, espa, and Vandādaʾi varf, pür, kuve; sentences: Latḥori, Ḵosr. to-rā did-am, Ḵonbi to-m be-di, Vandādaʾi to-m bi-di “I saw you”; Latḥori u rafte bu nū bexarra, Ḵosr. un rafte bud nun bexere, Ḵonbi nā bošta bu nū bereyin, Vandādaʾi un bišta bo nun birinu “he had gone to buy bread.” Accordingly, we may make the following observations. (1) The two Persian varieties, though from comparatively remote villages, are practically identical with one another and to the colloquial Persian of Tehran. This suggests a fairly recent language shift to Persian. (2) The Persian varieties contrast sharply with the Rāji dialects in vocabulary, morphosyntax, and historical phonology. (3) The Rāji dialects are essentially similar in morphology and syntax; for instance, past transitive verbs show ergative construction (-m bedi), and the past participle has a modal prefix (bo /bi-šta). (4) They differ in major lexical items (espa ~ kuve “dog,” u ~ “he”). As also shown in Isogloss no. 2, below, geographical proximity does not guarantee the isoglottic identity or likeness among the Kashan Rāji dialects. 

Abbreviations. Abu (Abuzaydābādi), Aby (Abyānaʾi), Ārā (Ārāni), Ard (Ardestāni), Bād (Bādrudi), Bdg (Bidgoli), Bid (Bidhandi), Bij (Bijgāni), Del (Delijāni), Far (Farizandi), Han (Hanjani), Jow (Jowšaqāni), JKāš (Jewish Kashani), Ḵān (Ḵᵛānsāri), Keš (Kešaʾi), Mah (Maḥallāti), Mey (Meymaʾi), Nat (Naṭanzi), Nar (Narāqi), Naš (Nešalji), Nāy (Nāʾini), Qāl (Qālhari), Qoh (Qohrudi), Soh (Sohi/Soʾi), Tār (Tāri), Trq (Ṭarqi), Vān (Vānešāni), Vār (Vārāni), Vrk (Varkāni), Yār (Yārandi), Zef (Zefraʾi), Zor (Zori). No abbreviations are used for the poorly documented dialects of Ḵonb in Markazi district, Kāḡazi in Kavirāt rural district, Niāsar, Azvār, Borzok, Vidjā, and Viduj in Niāsar rural district, and Totmāj and Javinān in Qohrud rural district. 

Comparative studies. For well over a century, the scholars of Iranian dialectology have made several attempts to classify Central dialects into meaningful groups, aside from geographical contiguity. These comparative studies became more refined as more dialects have been made known. Major studies are summarized in chronological order as follows. (1) In his vanguard study, Valentin Zhukovskiǐ (1888), one of the most prolific collectors of Central dialects, divided them into two groups: Kashani, consisting of Qoh, Keš, Vān, and Zef, and Isfahani, consisting of Gazi, Sedehi, and Kafrāni. (2) At the turn of the century, when sufficient information was available for a broad investigation of Iranian languages, Wilhelm Geiger defined Central dialects as a distinct group within West Iranian languages, and he further classified them into the subgroups of Kashani (Qoh, Nat, Vān, Mah, and Nāy), Keš–Zef, and Yazdi Zoroastrian. His criteria were fourteen linguistic features as well as geographical contiguity. (3) Karl Hadank (in Mann and Hadank) embarked on a detailed comparative analysis of Central dialects, but numerous overlaps of isoglosses prevented him from arriving at any concrete classification. (4) W. Ivanow’s classification is devoid of explicit linguistic analysis, though he seemingly considered sociolinguistics, which he introduced into Iranian dialectology, in defining a Kashan subgroup that consisted of JKāš, Qoh, Far, Yār, Keš, and Soh, while Nat is placed in Yazd subgroup and Ḵān, Vān, and Mah in Isfahan subgroup. (5) In an encyclopedia article, Harold Bailey (p. 1055) made a general survey of modern Iranian languages, and assigned Central dialects into five subgroups, which correspond closely to subsequent classifications, though his comparative principles are not explicit. Two of these subgroups include the dialects of Kashan: Yār, Far, and Nat versus Qoh, Mey, Soh, Keš, and Zef. (6) The next major classification of Iranian languages belongs to Georg Morgenstierne (1958, pp. 71 ff.), who founded his study on historical phonology and, to a lesser degree, morphology. He classifies more than eighty Iranian languages and dialects into twenty-three groups, eight of which concern Central dialects. Like Bailey, Morgenstierne assigns the Kashan dialects into two distinct groups, namely Yār, Far, Nat, and Ard versus Qoh, Keš, Soh, and Mey, while Zef is grouped with the dialects of Isfahan. (7) A significant step forward in the study of Central dialects was Karl Krahnke’s doctoral dissertation “Linguistic Relationships in Central Iran.” Of the twentyeight dialects he brought together in his study, twelve belong to the Kashan area: Abu, Qoh, Jow, Mey, Aby, Yār, Far, Soh, Bād, Nat, Keš, and Trq. His thirty-nine isoglottic maps are based on phonological and morphological-lexical features. The present comparative study owes much to Krahnke’s findings, but it refines and expands the latter by including more localities (JKāš, Ārā, Bdg, Del, Naš, Tār) and presenting new isoglosses that are chosen purposely to be relevant to the Kashan area. (8) A well-received classification of the Central dialects belongs to Pierre Lecoq (1989), who marks out the Kashan-Naṭanz area as forming the northeast group, while Ardestāni and Zefraʾi are placed in the southeast group together with Yazdi. (9) In an authoritative study, Gernot Windfuhr (1992; see central dialects) defines a large northern-central group in the area of Kashan and Naṭanz, which excludes Ardestāni and Zefraʾi, both assigned to the southern group. (10) Rastorgueva and Moshkalo conducted an extensive analysis of Central dialects, but without any explicit classification. (11) Donald Stilo (2007; see isfahan xx) groups the Central dialects areally into four geographical quadrants and defines the Isfahan dialects by drawing the bundles of eleven isoglosses (idem, fig. 5). His approach is taken as a model in our isoglottic analysis below. 

The data. The linguistic data used in the present study are obtained from the following sources: Zhukovskiǐ (JKāš, Qoh, Keš, Nat, Vān, Zef), Mann and Hadank (Nat, Soh, Mah, Nāy), Christensen (Far, Yār, Nat), Andreas (Soh), Lambton (Jow, Mey), Eilers (Ḵān), Krahnke (Bād, Trq), Lecoq, 2002 (Qoh, Abu, Aby, Tār, Bād), Yarshater, 1989 (Ārā, Bdg), idem, 1985 (grammatical gender in all dialects), Majidi (Mey, Vār, Bij, Nar, Naš, Qāl, Zor), Zargari (Jow), Ṣafari (Del), Āqā-Rabiʿ (Han), author’s field notes (Far, Ḵān, Keš, Nat, Trq, Zef), and the EIr. entries: abuzaydābādi; abyānaʾi; bādrudi; bidgol and bidgoli dialect; isfahan xxi. provincial dialects; jowšaqān ii; meyma ii (online); qohrud ii (forthcoming online).


Phonology. The consonant inventory of the Rāji dialects is essentially similar to Persian. Some dialects have the pharyngeal fricative /ḥ/ [ħ] and stop /ʿ/ [ʕ], but neither is phonemic. These sounds appear predominantly in Arabic words, hence are clearly influenced by Arabic; but no compelling explanation has been put forward as to how the dialects have adopted them. In their vocalic system, the dialects show little similarity; they range from as few as eight phonemes in Ārāni to fourteen in Abuzaydābādi. The highly developed vocalic system of Ṭāri is one of the richest in the world with twenty distinct vowels, if we include the five nasal vowel sounds for which Lecoq (2002, p. 22) establishes minimal pairs. Most dialects have the fronted ö and ü; the phonemic status of neither is easy to establish. Similarly, the vowel length recorded by the earlier collectors is seldom justified when examined at the phonemic level in the more fine-tuned, later studies. 

Nouns. Nominals are inflected primarily with ‑a and/or ‑e, which play several roles in each Rāji dialect. (1) Feminine marker is ‑a and ‑e (see Isogloss no. 2, below). (2) The indefinite is marked with an unstressed ‑e or ‑i: Abu (ī) pǘr-ē “a boy,” Aby (ya) rū-e “one day,” Mey jan-e “a woman.” (3) The definite marker is ‑a or -e (stressed, as in colloquial Persian): Qoh yene-y-a/e “the woman,” Abu pe-y-a “the father” (with epenthesis -y-), Aby kāši-a “the Kashani.” (4) The eżāfa marker (-e or -i), probably not genuine to the Kashan area, occurs occasionally, either explicitly (Mey bar-ε raz “garden’s door”) or by altering the final vowel: Qoh keye man “my house” (cf. keya “house”), Abu kēye ma “my house” (kēya “house”). In Abyānaʾi, the eżāfa may mark a modified feminine noun: dót-a man “my daughter,” sāl-a dövvömji “the second year” (Lecoq, 2002, pp. 60, 71). (5) The definite object marker is sometimes reduced to -a or ‑e (= Pers. -). (6) An unstressed -a or -e occurs occasionally on the nominal complement of verbs or direct objects in Bidgoli: kār-e xeyli dāro “I have much work [to do],” dars-a axune “she studies” (Yarshater, 1989, p. 381); a similar suffix in Abyānaʾi is interpreted as the optional feminine marker, e.g., zendegāni-a akerān “I live” (Lecoq, 2002, p. 60). (7) An optional unstressed particle, with no semantic value, exists in various dialects, e.g., Qoh mer/ méra/mére “man” (Lecoq, 2002, pp. 76-77). (8) The plural marker is -a in Abyānaʾi (e.g., meša “ewes”), -e in Bādrudi (bāle “spades”); other dialects employ -(h)ā or -un

In fact, the defining line between these functions is sometimes blurred, and their interference makes the morphology of the dialects difficult to understand. It may look even more complicated if we consider the role of the same suffixes in lexical distinction (Abu döt “daughter” ~ doča “girl,” Aby doté “girl” ~ doÅŃ ta “daughter”). Moreover, a whole class of words in each dialect carries the ending vowels /-a/, /-e/, or both, derived from the Old Iranian suffix *-aka (see meyma ii, for instance). This complex aspect in morphology calls for a more detailed study of individual dialects. 

Pronouns. Concerning the freestanding first singular pronoun, the historical direct/oblique binary is preserved only in Aby az/man, Bād a/men (see bādrudi; cf. Bād axo / min in Lecoq, 2002, pp. 83, 278); other Central dialects have adopted the undifferentiated Persian form. The third singular personal and demonstrative pronouns with nasal initial are characteristically Kashani (see Isogloss no. 3, below). Some dialects show distinct forms for gender (masc./fem.): remote Abu na/nön, Aby nūn/nūna, Jow nun/núna, Naš nu/nemun, Del on/ona; proximate Abu / nēm, Aby nēn/nēna, Jow nen/néna, Naš na/nuhun, Del in/ ina (see also Isogloss no. 2). 

Enclitic pronouns (Table 2) are generally similar to those of Persian, but note these phonological changes: (1) The original consonant /š/ in the third person singular and plural is /y/ in Ārāni, Abuzaydābādi, and Qohrudi. (Sedehi and Sagzavi too have 3rd sg. -y, while other provincial dialects of Isfahan have either -ž or -š.) The arbitrary distribution of the types in /š/, /ž/, and /y/ among the Central dialects suggests the historical line of change š > ž > y, in which the original š is first voiced into ž, and the letter is lateralized to become y. This author has observed this diachronic development as being currently active among different generations of the speakers of Ḵᵛorzuqi and Gazi (see jazi). A similar trend may be seen in šm > žm > ym, as shown below, under Isogloss no. 1, as well as in j > ž > y in “woman.” (For an alterative explanation, which assumes that the third singular -y and -š are derived respectively from OIran. *hai and *šai, etc., see Windfuhr, 1975, 1992.) (2) An original /t/ or /d/ in the second person singular and plural is softened to /y/ in Farizandi and Yārandi. Interestingly, Abyānaʾi, located in a buffer zone between these two groups, shows one aspect of each: 3rd sg. -y and 2nd pl. -yi (< yü < yũ < yun). Subsequently, ketāb-ey means “his book” in Abyānaʾi but “your book” in Yārandi; the two neighboring vernaculars are just a few kilometers apart in the Barzrud mountain valley. 

Verbs. The significant role of transitivity, combined with the secondary construction of most past stems (by suffixing -ā to the present stem), has led Lecoq (2002, p. 119) to divide the past stems into four classes; examples of present : past stems in Qohrudi are: (i) transitive, ending in a consonant, ger- : gera(t)- “seize”; (ii) transitive, ending in a vowel, geln- : gelnā- “rotate”; (iii) intransitive, ending in a consonant, k- : kat- “fall”; and (iv) intransitive, ending in a vowel, gīn- : gīnā- “rotate.” Some of the dialects are capable of forming an inflectional passive (see Isogloss no. 5, below), and all form causative stems by suffixing -(V) n to the present stem, e.g., Ārā -m-tej-ən-ād-a “I have caused to run.” 

Personal endings (Table 3). The first person singular verb ending is -on, -õ, -o, or ‑un in all Kashani dialects, except JKāš -om, Jow -am, which are seemingly formed under Persian influence. The third singular is typically -a or -e; the clearly different -u in the Jewish dialect of Kashan suggests influence from Isfahan provincial dialects. Some dialects show a different ending for the feminine third sin- gular: Abu, Aby -a, Qoh -e, Del -ae, Jow -ea. The ending is zero in the preterit, but Jow has also -e. See also Isoglosses nos. 2, 4, 6, 8, below. 


The fourteen features that follow are chosen to define the Kashan area within the entire region of Central dialects. Selection of the isoglosses is based on the availability of data for most dialects as well as their specificity to the Kashan area; scarcity of data and overlap of isoglosses prevented us from including some of the features identified in previous studies. Our isoglosses can be divided into phonological (1), morphological (2-6), and lexical (7-14) features.  For succinctness, when a word is given in small-capital letters, this indicates that a range of forms occur in different dialects. Some of the isoglosses are shown in Table 5

1. “Eye.” The gloss “eye” (cf. Av. čašman-, MPers. and NPers. čašm, colloquial Pers. češ) is represented by two forms in Central dialects: the ca(y)m type (e.g., Borzoki, Vrk čaym, Han čam, Yār čem) prevails among the Kashan and other northern dialects versus caš in the south; the bisecting line passes through Zefra, for which both forms are reported: čem (Zhukovskiǐ) and češ (my data; see also Krahnke, no. 4). A parallel reduction of the original -šNcluster is found in more words: Qoh paym, Jow paim, Nat poime (cf. Tār pežm, Keš pajm, Aby pāzm) “wool”; Aby tena, Jow taina “thirsty”; Mey powna, Keš pōina “heel”; Aby enoyn-, Jow einās- “hear, recognize” (< OIr. *xšnā-sa-). Although sufficient data is lacking in other dialects for a wide-ranging comparison, the accessible data suggest the line of sound change -šm- > žm > (y)m (cf. Windfuhr, 1975). For a related development, see above, under Pronouns. 

2. Gender distinguished. Several Kashan dialects have grammatical gender in various domains, as shown in Table 4 (summarized from Yarshater, 1985, with addition of Farizandi). The feminine marker for nouns, as well as for pronouns and adjectives, is an unstressed -a, e.g., Del, Aby bā́ la “spade,” Del béza, Naš bóz(a) “goat,” while Jowšaqāni has the feminine in -e and the masculine in -a, as in boze “goat” ~ varzāa “ox”; varge “she-wolf” ~ varga “he-wolf.” In the dialects which show no formal gender marker in nouns, the distinction is revealed by accord of adjectives and verb. Examples: (adjectives) Abu böz esbíde “white goat,” Aby sǘra goÅŃ la “red flower”; (indefinite articles or numerical adjectives) Del ajey pür “a son” ~ ija dete “a daughter,” Aby e dāde “a brother” ~ ya dādā “a sister”; (demonstrative adjectives) Del on purá “that boy” ~ óna zenčo “that woman”; for demonstrative and personal pronouns, see Pronouns, above. 

Gender is better preserved in the verb phrase as an integral part of the inflectional system. Examples are: Naš pür/ dot-am nāsāz-e/a “my son/daughter is ill”; Qoh mer bömö “the man went” ~ yan bömöde “the woman went”; Del berā-m vāgardā be “my brother had returned” ~ fāka-m vāgardo-wda “my sister had returned,” Naš berā/xāka Hasan-am bedi bö/be “I had seen Hasan’s brother/sister” (for constraints, see Yarshater, 1985; Lecoq, 2002, pp. 58 ff.; see also central dialects). Farizandi seems to mark gender only in some third person verb agreements when the subject is naturally feminine, e.g., pür/dot-äm ko-šo / štä? “where did my son/daughter go?” As demonstrated by Yarshater (1985, pp. 741-43), Meymaʾi and Ardestāni also show vestigial remnants of gender in the verb phrase.  

A closer examination of the Kashan dialects without the category of gender reveals that some of their nouns carry the unstressed ending -a or -e that can be traced back to the Old Iranian feminine marker *-ā (rather than *-aka, the source of -a/-e in Qoh espá, JKāš esbé < *spaka- “dog,” and most other nouns in West Iranian that end in these vowels). Morgenstierne surveyed this category of nouns in Tati and Central dialects; his examples from Kashan area include (expanded here): Soh bǘzä, Bij, Vār beza “goat” (cf. Del, Naš, Abu, above); Soh kaÅŃ rge, Qoh kárg(e), Nat kárga, Bij, Vār kerga (cf. Naš and Aby fem. kárga) “hen”; Qāl bara (cf. Abu fem. bar) “door”; Mey düme “picture” (cf. Aby fem. díma “face”); Nat dúta, Qoh, Keš, Zef dúte, Vān déte “daughter.” Morgenstierne’s investigation suggests that gender distinction, at least in the nominals, is an archaic feature which once existed in all Kashan dialects as well as in Tati in the north. 

This notion accords well with the findings in an areal study of Donald Stilo (unpublished paper). He shows that, in terms of grammatical gender, the Kashan dialects form the southern end of a larger language continuum that extends with little interruption northward to Āmoraʾi, Vafsi, and Alviri, then to Southern Tati of Qazvin and Upper Ṭārom, and farther north up to Kajali. Moreover, Stilo puts forward the idea that certain Tati dialects have not only been conservative in retaining an older gender distinction but have expanded it into new domains through cross-referencing the gender of the noun in different loci within the noun or verb phrase. Focusing on the neighboring Tati dialects of Kafteji and Kelasi in eastern Ṭārom, Stilo then explains how these two closest genealogical relatives have taken opposite direction: the latter having lost gender completely, while the former is one of the richest among Tati dialects in terms of gender. Stilo’s findings can be applied to similar situations in the Kashan area, namely Jowšaqāni and Kāmuʾi, Qohrudi and Javināni, and Ārāni and Bidgoli—three pairs of contiguous dialects that have taken opposite paths in gender. 

3. Third singular pronoun. The third singular personal and demonstrative pronouns in the Kashan dialects is the nasal initial no(n), nu(n), or the like (Krahnke, no. 21), which contrasts with most other Central dialects in the west and south which share the vowel initial forms also found in Persian. Although without established etymology, the Kashani pronominal forms must be old in that they differentiate gender in the dialects with gender distinction. 

4. Durative marker. To mark the present indicative and the imperfect, four major strategies are employed in Central dialects: the prefix et-/at- in the western area from Maḥallāt to Ḵᵛānsār, the prefix e-/a- in the Kashan area, the suffix -e in the Isfahan area, and no marker in the rest of the dialects, including Naṭanzi (see Krahnke, pp. 182-87; Lecoq, 2002, pp. 110 ff.; Windfuhr, 1992, pp. 249-50). In a recent study Stilo (2007, pp. 106 f.) has established that the Kashani durative marker e-/a- originated from an older et-/at- (as in the dialects to its west), on the ground that the latter forms appear before the stems that begin with vowels, e.g., Aby e-kar-ān “I do” ~ et-özmar-ān “I count.” It should be added here that (1) the use of the t- variants is not consistent before vowelinitial stems; it is rather limited by and large to two verbs “bring” and “come”: Abu at-ār-ō “I bring,” ma-t-avar “I would bring,” a-t-ō “I come,” at-amd-ō “I would come”; Qoh at-ār-ūn “I bring,” at-ömöd-ūn “I come”; Aby et-orān “I bring,” me-tt-ārd “I would bring”; Tār at-ār-õ “I bring,” am-t-ā “I would bring,” at-õmo-yõ “I would come”; Mey at-ema-iy-ɔn “I would come.” (2) For these verbs, the historical t- has become the frozen part of the stems in Meymaʾi and Jowšaqāni, as they appear in non-durative forms: Jow na-t-ār “don’t bring!” ba-t-am “(that) I come” (subjunctive); Mey betā “bring!” -tɔr-e “that he bring,” be-šun-“they brought,” eidi-aš be-tɔrde “he has brought presents.” These lexicalized forms with the initial t- have misled Cheung (p. 9) in proposing the root *tar “to cross over” for Jow, Mey tār- : tārd- “bring.” This verb is indeed derived from Old Iranian *ā-bara- “bring,” which, like *āy- : *ā-gmata- “come,” had the long initial vowel ā, and this may have acted as the catalyst in preserving the full form of the later prefix *at-. 

5. Inflectional passive. Most Kashani dialects build a passive/intransitive form with the infix -i- (< Mid. Ir. -ī(h)-), placed after the present stem. This infix is usually realized as -iy- (with the connecting glide y) when an ending or the past-stem marker ‑ā(d)- is added. Qohrudi and Abyānaʾi employ the infix -(e)g-, probably the palatalized form of an original -y-, and Ḵᵛānsāri has -k-. Examples: JKāš āssin-um viar-i-ad-e “my sleeve is torn,” Han a-rij-i-a “it pours,” a-rij-i-ā “it was pouring,” Bidhandi rij-i-ā “it poured,” Trq be-hmir-i-ā “it broke,” Mey -š behmeriyo (i.e., be-hmer-i-ā) “his foot got broken, broke,” Aby ba-hmar-g-ā “it broke,” Qoh a-kod-eg-ūn “I get hit,” ba-rij-ig-ād-e “it was poured,” Ḵān ba-xor-k-ā “it was eaten.” Bidgoli shows the sole example bahmeyrā “it broke,” seemingly from *ba-hmar-i-ā. The passive and causative infixes may coexist; e.g., Jowšaqāni stems voz- : vašt- “run” yield ba-vašt-ø “he ran,” ba-š-voz-n-ā “he made [it] run,” ba-voz-n-i-ā-ø “it was made run.” 

6. Future with kam-. An important feature that defines the Kashan area from the rest of the Central dialects is the periphrastic future tense auxiliary: JKāš, Qoh, Han, Abu kem-, Ārā kəm-, Far, Soh kam-, Aby kām-/köm-, Yār, Bād kom- (cf. Mid. Pers. kām- “wish”). This auxiliary is followed by the past stem, i.e., the “short” infinitive: Ārā kəm-o /-e/-e šu “I/you/he will go,” Bād kom-ūn/i/ä kard “I, you, he will do,” Han kem-un yošd “I shall find,” kem-a gā “he will want,” kem-a vijiā (passive) “it will be said.” The same structure is found in the Persian variety of Kashan with the auxiliary verb ḵᵛāstan (e.g., -m xarid “I will buy”; cf. formal Pers. ḵᵛāh-am ḵarid) but is absent in colloquial Tehrani and Isfahani Persian. In Jowšaqāni and Meymaʾi a future is formed by the invariable komi preceding conjugated forms of the main verb, e.g., Jow komirāst- ima “we will weave,” dar-komi-pušt-inda “they will put on” (cf. Aby dar-kəm-im-ka “we shall fall”). 

7. “Now.” The form hat spreads all over northern area of Central dialects, sharply contrasting with the forms zonon, zogon, and osme in the south. This isogloss is often used in delineation of Central dialects. See Krahnke, no. 30; Stilo, 2007, fig. 3. 

8. “Sit.” This feature is well discussed by Stilo, who shows the distribution of two distinctive forms in Central dialects: hā-čin- : hā-češt in the northwest, including Kashan, and hā-ni(n)(g/k)- : hā-ništ in the southeast, with Naṭanz and its villages Ṭār and Keša forming a buffer zone where the two isoglosses meet (Stilo, 2007, fig. 6). Yet a third set of stems can be identified in Central dialects for “be seated, remain”: Mey āh- : āhast-/āhāst-, Qoh āh- : āhā(d)-, Tār ax- : axā(y)-, Ard ax- : āxo-, Aby ay- : ayā(y)-, Abu av- : avad-, Nāy, Anāraki - : āšes(s)- (Lecoq, 2002, p. 194). Cheung (pp. 126, 154) assumes that these stems have derived from the root *had (prefixed with *ā-). 

9. “Such.” For the gloss “such, like this,” corresponding to Pers. čonin, three types prevail in the Central dialect area: in most of the northern part nezan (e.g., Han nezen, Jow nizan, Mey εzina) and to its south and east issin and sige (see Stilo, 2007, fig. 4). But note also Del isin, JKāš engi (< ēn gōna?). 

10. “Big.” This isogloss dissects the Central dialects neatly into four geographical quadrants, with Kashani (northeast) gord, southwest bela, southeast and northwest mas(sa). Interestingly, the Persian variety of Kashan has retained (or borrowed) this gloss as gurd. See Krahnke, no. 28; Stilo, 2007, Tab. 2, Fig. 7, and p. 108. 

11. “Small.” Contrary to the prevailing form kas “small, little” in the rest of Central dialect area (Stilo, 2007, table 2), a group of contiguous locations in Kashan area employ a variety of forms, like vüjüj, that seem to have derived from a common ideophonic origin. 

12. “Wall.” Three distinct lexical forms exist for wall: kal in Kashan area, čina to its immediate southeast, and dezār and divāl elsewhere (cf. Krahnke, no. 3). 

13. “Sister.” The prevailing forms for “sister” and “brother” in the Kashan area are dādā and dāde, respectively, contrasting with the south and southeast Central dialects, which carry forms cognate with Pers. ḵᵛāhar and barādar (Krahnke, no. 34). The lexical pair dādā and dāde seems to provide an ideal pattern of gender distinction in an earlier stage of these dialects, with a remnant of the Old Iranian feminine ending *-ā (see Isogloss no. 2). There is yet another form for “sister” in the dialects to the west of Kashan: Nar, Bij fāk, Del fā́ ke, Mah, Vār xāk. The change of the initial *hw- to f is also found in more words, such as Del fās- : fāšt- “ask in marriage,” Farang (toponym near Ḵomeyn; < Xwarnaq; Yarshater, 1985), an isogloss that unites the area with Ḵuri (q.v.) and Sivandi (see central dialects). 

14. “Tomorrow.” This is a rather broad isogloss with predominantly two major forms among Central dialects: haya and sobā. The first form (from *fradā < *fratāka-), found only in Kashan area, represent two significant sound changes: fr- > h(r) and -d- > y (cf. Krahnke, no. 9).

Isogloss bundles. Table 6 and Figure 1 bring together the fourteen isoglosses presented above. The list includes twenty-three localities for which sufficient evidence is on hand. These belong to Kashan proper and its neighbors; Delijān, Maḥallāt, Vānešān, and Ḵᵛānsār on the west, Gaz and Zefra on the south, and Ardestān and Nāʾin on the east of the Kashan area are included for the sake of comparison. Paucity of data did not allow putting some important localities, such as Nešalj, Qālhar, Varkān (between Kashan and Delijān), and Ārān on the map. 

As seen in Table 6, Abyānaʾi ranks first by meeting all the fourteen isoglosses that set the Kashan linguistic area apart from its neighbors. At the bottom of the table, we find Nāʾini, which shows zero isoglottic agreement. When a feature is unknown in a dialect, the isogloss number is placed in parentheses with a question mark. For instance, Farizandi is obscure in Isogloss no. 9; it has therefore only thirteen known features to examine. Of these, twelve features meet the criteria set for the Kashan area. The feature that fails to pass is no. 11, glossing “small,” for which Farizandi kasl does not show affinity to the isoglottic form vüjüj. Having matched twelve out of thirteen of the known isoglosses, a ratio that yields 92 percent, Farizandi ranks fifth among the dialects included in the table. In Figure 1, therefore, Farizand falls within the bundle of 80-100 percent of the isoglosses. As we move away from this core bundle and cross over successive contour lines, the relatedness of the dialects to those in the core decreases increasingly. 

Figure 1 shows the core isogloss bundle as a triangle centered at the inner Kargas settlements of Qohrud, Abyāna, Farizand, Yārand, and Jowšaqān, with a corner reaching the garmsir Abuzaydābād on the edge of the desert. This implies that climate variation plays no essential role in defining the Kashan linguistic zone. Interestingly, however, Kashan itself, represented by its Jewish dialect, and its neighbor Bidgol, both on the garmsir plain, find their place on the map two bundles away from the core. Within the same bundle lies the relatively far-off Delijān as well as Vānešān in the valley of Ḵᵛānsār (q.v.), pointing to the fact that the Kashan area ranges itself more closely with its western neighbors than with Isfahan in the south. On the other hand, Naṭanz and its villages yield somewhat low isogloss ratios in view of them often being grouped together with the Kashan dialects (see Comparative studies, above). 

Lastly, Gazi and Zefra’i in the south and Nāʾini, beyond the frame of Figure 1, in the east have low degrees of relatedness to the Kashan dialects by showing little isoglottic agreement, and we should expect other Central dialects to be within the same weakest bundle on the map. It appears that the zone within the bundle of 40 percent or more in Figure 1 accords well with the region where the Esbandi festival is celebrated (see above, vi). Another sociolinguistic link can be between the position of Jowšaqāni within the core bundle of the Kashan dialects and the strong sense of belonging that the residents of Jowšaqān show toward Kashan rather than Isfahan. 


F. C. Andreas, Iranische Dialektaufzeichnungen aus dem Nachlass, ed. K. Barr, W. B. Henning, and A. Christensen, Berlin, 1939.

Abu’l-Ḥasan Āqā-Rabiʿ, Guyeš-e rāji-e Hanjan, Tehran, 2004.

H. W. Bailey, “Persia. II. Language and Dialects,” in EI¹ III, London, 1934, pp. 1050-58.

H. Borjian, “Median Succumbs to Persian after Three Millennia of Coexistence: Language Shift in the Central Iranian Plateau,” Journal of Persianate Societies 2/1, 2009, pp. 62-87.

Idem, “The Dialect of Jowshaqan, Part I” Iran and the Caucasus 14/1, 2010, pp. 83-116.

Johnny Cheung, Etymological Dictionary of the Iranian Verb, Leiden and Boston, 2007.

Arthur Christensen, Contribution à la dialectologie iranienne [I.] Dialecte guiläkī de Recht, dialectes de Färizänd, de Yaran et de Natanz, avec un supplément contenant quelques texts dans le persan vulgaire de Téhéran, Copenhagen, 1930.

Wilhelm Eilers and Ulrich Schapka, Westiranische Mundarten aus der Sammlung W. Eilers II. Die Mundart von Chunsar, Wiesbaden, 1976.

Wilhelm Geiger, “Centrale Dialekte,” in W. Geiger and E. Kuhn, eds., Grundriss der iranischen Philologie I/2, Strassburg, 1895, pp. 381-406.

W. Ivanow, “The Gabri Dialect,” Rivista degli studi orientali 16, 1935, pp. 31-97; 17, 1937, pp. 1-39; 19, 1939, pp. 1-58.

Karl John Krahnke, Linguistic Relationships in Central Iran, unpub. diss., University of Michigan, 1976.

Ann K. S. Lambton, Three Persian Dialects, London, 1938.

Pierre Lecoq, “Les dialects du centre de l’Iran,” in Rüdiger Schmitt, ed., Compendium linguarum Iranicarum, Wiesbaden, 1989, pp. 313-26.

Idem, Recherches sur les dialectes kermaniens (Iran central), Acta Iranica 39, Leuven, 2002.

Oskar Mann and Karl Hadank, Kurdisch-persische Forschungen III/1. Die Mundarten von Khunsâr, Mahallât, Natänz, Nâyin, Sämnân, Sîvänd und Sô-Kohrûd, Leipzig, 1926.

Moḥammad- Reżā Majidi, Guyešhā-ye pirāmun-e Kāšān o Maḥallāt, Tehran, 1975.

Georg Morgenstierne, “Neu-iranische Sprachen,” in B. Spuler, ed., Handbuch der Orientalistik I/4/1, Leiden, 1958.

Idem, “Feminine nouns in -a in Western Iranian dialects,” in W. Henning and E. Yarshater, eds., A Locust’s Leg, London, 1962, pp. 203-8.

M. Purriāḥi et al., Šenāsāyi-e guyešhā-ye Irān I-IV, Tehran, 1983- 2000.

V. S. Rastorgueva and V. V . Moshkalo, “Dialekty tsentralnogo Irana,” in Osnovy iranskogo yazykoznaniya. Novoiranskie yazyki: sever-zapadnaya gruppa II, Moscow, 1997, pp. 195-329.

Ḥosayn Ṣafari, Vāža-nāma-ye rāji: guyeš-e Delijān, Tehran, 1994.

Donald Stilo, “Isfahan xxi. Provincial Dialects” in EIr. XIV/1, 2007, pp. 93-112.

Idem, “Loss vs. Expansion of Gender in Tatic Languages: Kafteji (Kabatei) and Kelasi,” unpublished paper.

Gernot Windfuhr, “Isoglosses: A Sketch of Persians and Parthians, Kurds and Medes,” Monumentum of H. S. Nyberg II, Acta Iranica 5, Tehran and Liège, 1975, pp. 457-72.

Idem, “Central Dialects,” in EIr. V/3, 1992, pp. 242-52.

Ehsan Yarshater, “Distinction of Grammatical Gender in the Dialects of Kashan Province and Adjoining Areas,” in Papers in Honour of Professor Mary Boyce II, Acta Iranica 25, Tehran and Liège, 1985, pp. 727- 45.

Idem, “The Dialect of Ārān and Bidgol,” in C. H. de Fouchecour and Ph. Gignoux, eds., Études Irano- Aryennes offertes a Gilbert Lazard, Paris, 1989, pp. 371-83.

Raḥmatallāh Zargari, Tāriḵ, farhang, honar-e Jowšaqān-e qāli, Tehran, 1995.

Valentin A. Zhukovskiĭ, Materialy dlya izucheniya persidskikh narechiĭ, 3 vols., St. Petersburg, 1888-1922.

(Habib Borjian)

Originally Published: May 1, 2012

Last Updated: May 14, 2012

This article is available in print.
Vol. XVI, Fasc. 1, p. 38-47

Cite this entry:

Habib Borjian, “KASHAN ix. THE MEDIAN DIALECTS OF KASHAN,” Encyclopaedia Iranica, 2012, XVI/1, pp. 38-47, available online at http://www.iranicaonline.org/articles/kashan-ix-the-median-dialects-of-kashan (accessed online on 30 December 2012).