ZOROASTRIANS OF IRAN vi. Linguistic Documentation



vi. Linguistic Documentation

This article focuses on the importance of documenting the Zoroastrian dialects of Yazd and Kerman, also known as Zoroastrian Dari (a term not to be confused with classical Persian Dari or Dari in Afghanistan). We discuss the relevance of social, historical, and cultural contexts in the study of Zoroastrian Dari and the role of these contexts in the grammatical analysis of the recorded materials. For successful and productive documentation, we must pay attention to many factors both preceding and following the actual tasks of language documentation. J. H. Hill, for example, states, “If we are to succeed in sensitive documentation, which by definition requires the deep involvement of communities, we must incorporate a cultural and ethnographic understanding of language into the very foundations of our research” (Hill, p. 113).

Let us first look at Zoroastrian Dari and review the importance of its documentation and its study to different fields. Dari is an endangered language spoken by the Zoroastrian minority, who mostly live in Yazd and the surrounding areas, and in Kerman and Tehran. Zoroastrian Dari is one of the most interesting Iranian languages for various reasons. One reason is that Dari is spoken only by members of a religious minority, the Zoroastrian community, who use it as a secret language. According to Hill, “One of the reasons history speeds up at the margins is that oppression and marginalization—and minority and indigenous language communities are almost by definition oppressed and marginalized—produces a special intensification of language-ideological projects” (Hill, p. 114). This observation is absolutely applicable to the case of language use in the Zoroastrian community.

The study of Zoroastrian Dari is of particular importance for Iranian dialectology and comparative linguistics. This language is used in a parallel way to the Persian language of the Muslim population, and one can observe strong influence from Persian, especially in the domain of the lexicon. But Dari also differs from Persian, having special characteristics common to the languages of the Northwest Iranian group. Dari is also similar to the Jewish dialect of Persian in Yazd and Kerman (see Yazd iv. The Jewish dialect of Yazd). T. E. Gindin mentions that Dari in Yazd shows some similarities to Judeo-Yazdi, but points out that it also contains differences that affect language use to the point where speakers cannot understand each other. It is noteworthy that, particularly in Yazd, there is no social communication between the Zoroastrian and Jewish communities. The present article does not treat the comparison of Dari and Jewish dialects in further detail. The language of Dari has not yet been sufficiently studied, and for this reason, there is still much to research with regard to the different aspects of this language. All Dari sub-dialects are seriously endangered, and it is necessary to start their documentation as soon as possible, because the study of each sub-dialect contributes material of particular importance to a description of the entire language group.

Up to the present day, the question of the degree of linguistic distance of the Kermani dialect from other Yazdi dialects has not been answered. Through my project, “Documenting a religious minority: Zoroastrian Dari in Kerman,” funded by the Endangered Languages Documentation Programme, at the School of Oriental and African Studies, University of London, I have been documenting Zoroastrian Dari in Kerman. In order to clarify the relationship between the Kermani and Yazdi dialects, I have also recorded two further Yazdi sub-dialects, Qāsemābādi and Elābādi, compared selected features, and described their similarities and differences.  There are many other Yazdi sub-dialects that have not yet been documented or systematically studied, and these too must be documented.

Before starting a systematic documentation project, it is necessary to take some relevant factors into account. These factors are discussed in the next sections.


Zoroastrian Dari is a critically endangered language. The migration of Zoroastrian families to the big cities, and the mass migration of young Zoroastrians, especially to the United States and to Canada, and their social integration in the new homeland, is the main reason for the high endangerment status of this language. If we look at the number of Zoroastrians in 1879 in Iran, provided by A. Houtum-Schindler (p. 55) and compare it with the last official set of statistics from the Statistical Center of Iran in 2011, we observe that the number of Zoroastrians declined in Kerman and Yazd during 1879-2011, while it increased by more than 5,745.33 percent in Tehran during the same period. This change in numbers is evidence to the mass migration of Zoroastrians to Tehran (see Table 1, Table 2, Table 3). In Tehran, only a limited number of Zoroastrians, mostly elderly people, still speak Dari. Among members of the younger generations, use of the language has been lost.

Migration is not the only reason behind the critically endangered status of Dari. Persian, the official language of Iran, has been strongly influencing all regional and minority languages, including Zoroastrian Dari. In terms of language endangerment, both Kermani and Yazdi Dari are regarded as critically endangered. The situation of the Kermani dialect is especially grave, while the situation of the Yazdi dialect of Dari is comparatively better. Today we know that only three elderly women are still proficient in Kermani Dari. All of them live in Kerman. Unfortunately, one of our best informants has become ill and has lost her ability to recall the language.

A question arises here regarding the reasons that Yazdi Dari has been better preserved than Kermani. There are various reasons for the difference in degrees of the preservation of Yazdi Dari. Yazd has been known as a conservative city in Iran in many respects. It seems that there has not been a close relationship between the Muslim and Zoroastrian communities during the history of this city. Zoroastrian Dari in Yazd can be characterized as isolated islands with its own rituals and sub-dialects, existing within a Muslim society. I have very often experienced a situation in which the speaker of Dari in a particular location perceives his or her dialect as the best and most unique one of all and is extremely proud of it. In my view, this positive attitude has contributed to the preservation of the individual sub-dialects in Yazd. Another reason for the preservation of Yazdi Dari has to do with the limited travel between villages. It is not common in Yazd that a Zoroastrian from one village will attend a ceremony in another village. When I asked one of my informants in Qāsemābād, a village presently within the city limits of Yazd, about any differences to be found between the Gahanbār ceremony (see Gāhānbār) in Qāsemābād and that in other villages, he answered: “I don’t know. I have never gone to a ceremony outside of Qāsemābād. We don’t usually do that.”

The difference in the degrees of preservation of Yazdi compared to Kermani can also be attributed to openness to the larger society. In comparison to the Zoroastrian community in Yazd, the Zoroastrian community in Kerman has been more open to the non-Zoroastrian society around it. It has been strongly influenced in many cultural and linguistic aspects by the predominant culture and language of Persian. The Zoroastrian community in Kerman has not been an isolated one, but has been more fully integrated with the non-Zoroastrian society. The Zoroastrian community there has not preserved the observance of many rituals or the use of their particular dialect to the degree that the Zoroastrians in Yazd have.

The special character of the Zoroastrian community in Kerman and the low numbers of speakers of its dialect create a very challenging and complex situation for fieldwork. One of the challenges has been to involve a significant number of people in the interviewing process, the basic method by which we obtain the language material necessary for documentation. For example, in Kerman, it has been our experience that when requested to give an interview, some people were not ready to participate without other speakers present. People were only convinced to participate if they could invite other speakers, most of whom are also related to one other. We did an interview with one person who was ready to participate, and it was only after that interview that other people began to speak. Unfortunately, many people are not able to speak easily and at length, particularly when they see a camera and microphone. We attempt to discuss the interview process with them and convince them by various means to speak as much as possible. The role of my colleague, Armita Farahmand, as a member and speaker in this community, is thus very important. Minu Mehrabāni is another member of the community and colleague in this project, who helped us find one of the oldest informants in Kerman and to do an interview with her.

The informants in the language community prefer to speak about special topics. These topics include religious ceremonies, proverbs and other relevant oral traditions, memories about childhood and school years, and details of preparation and cooking of traditional foods and sweets for the religious rituals. Women enjoy speaking about such topics and are not afraid of discussing them. Before beginning the interviews, I provide the speakers with a list of questions, which also help me to focus on special characteristics of the language. Typically, speakers talk to a great degree about their past and present experiences. For obtaining more material on a wider variety of tense-aspect forms (for example, subjunctive of the present, past, and future, as well as of the continuous aspect in present and past), we ask some more specialized questions depending on the situation and background of each speaker. A real challenge during my documentation work in October 2012 was the investigation of a possible infinitive form. Speakers did not use it in their conversations and did not know what an infinitive was. I had to use the infinitive forms in a question, to which the answer could be given with an infinitive form.

As previously mentioned, there are various relevant Yazdi sub-dialects. These sub-dialects are of particular importance for the study of Zoroastrian Dari and must be documented systematically. As the Zoroastrian community in Yazd is more close-knit and more conservative than that in Kerman, the situation for fieldwork is more complicated and challenging. The first factor to attend to before starting the fieldwork in Yazd is to study in detail the cultural and social characteristics of the Zoroastrian community in Yazd and the surrounding areas. The best situation would be for the members of the language community itself to engage in the fieldwork, while accompanied by a documentary linguist. Collaboration between community members and linguists in fieldwork would be ideal, but it is, in practice, difficult because of the sensitivity of the language community.


While the knowledge of the cultural and social context of the Dari language is a relevant factor in starting fieldwork and language documentation, the knowledge of the history of the Zoroastrian community and the Dari language is an essential requirement for the analysis of language materials and the study of Dari. The history of the Dari language is still unclear, as many historical aspects of the Zoroastrian community have not been sufficiently studied. In order to better understand Zoroastrian Dari, it is necessary to study the history of the Zoroastrian community in Iran. Zoroastrian Dari is the language of immigrants in Yazd and Kerman, and it is of particular importance to clarify the history of this immigration and the origin of these immigrants. We can assume that the speakers of this language originally came from various regions. The original languages of these new immigrants in Yazd and Kerman were probably from different language groups. The migration from east and west, north and south to Yazd and Kerman was a gradual process, and it is possible that the formation of this language as an independent language of the Zoroastrians occurred over a long period of time.

Let us look at the history of Zoroastrians after the Arab-Islamic conquest of the Sasanian empire (see Zoroastrianism ii. Historical Review: from the Arab Conquest to Modern Times). There is a big gap regarding the history of Zoroastrians in Iran after the Arab conquest in the middle of seventh century until the Safavid era from the sixteenth to the eighteenth century. It must be kept in mind that the Islamization of the country was not an immediate change, but rather a transition that took several centuries. Some historical books from the earliest centuries after the Arab invasion and the advent of Islam reported details regarding this process. For example, the Tāriḵ-e Sistān reports that many Zoroastrians had converted to Islam in the year 666 CE (p. 91), and Zoroastrian fire temples were destroyed in the year 671 (p. 93). But complete control over Iran probably happened after some centuries, around 1300 (Choksy, p. 143), that is, during the time of the Mongols. During this long period, “Zoroastrianism shrank into the role of a religious minority” (Stausberg, p. 175).

The immigration of Zoroastrians to Yazd and Kerman was a gradual process, as mentioned above. Before the Mongol invasion of Iran in 1219-24 CE, there must have been some Zoroastrian communities in different regions, especially in Khorasan and Sistān. But after that time, in the sixteenth century, the location of these communities was gradually reduced to two regions, Kerman and Yazd, with their surrounding areas. Two key questions that arise here involve the time of the immigration and the cities from which the Zoroastrians moved to Yazd and Kerman. Various scholars, such as C. E. Bosworth (p. 39) and J. Kestenberg Amighi (1990, p. 71), have mentioned a connection between the Mongol and Timurid invasion with the decline and immigration of Zoroastrians. In the late fourteenth century, the Khorasan area was particularly hard hit by the attacks of Timur. At that time because of this disaster in Khorasan, Zoroastrians moved to Sistān (Sorušiān, 1991, p. 17) and then to Kerman.

Two Zoroastrian documents provide evidence to prove the theory that the origin of Zoroastrians in Kerman is from Khorasan. The first document is Kāmdin Šāpur’s letter from 1559 CE. At the end of this letter we find a list of Behdins of Khorasan, who are in Kerman, with behdin meaning the people of good religion and referring to Zoroastrians. After the list, this information is mentioned: “The congregation of the Behdins of the country of Khorasan, who live here, is 3000 persons” (Dhabhar, p. 620). The second document is the letter of Frēdōn, the son of Marzbān, who states in his letter to India that his family, that is, the Marzbān family, a famous Zoroastrian scribe family in Kerman, originally comes from Khorasan province (Unvala, p. 153).

In contrast to the clear evidence for the origin of Zoroastrians in Kerman, there is less historical evidence regarding the origin of Zoroastrians in Yazd. For example, one of my informants, Mr. Bāstāni, who originally comes from Elābād but currently lives in Yazd, told me that his father says that the origin of their family is Isfahan. But we know that the migration of Zoroastrians from Isfahan to Yazd happened during the Safavid period. The general scholarly consensus is that the Zoroastrians of Yazd originally came from the western provinces of Iran, while the Zoroastrians of Kerman came from the eastern provinces. This question of place of origin is a relevant topic and needs to be sufficiently researched. This topic is of particular importance for its relevance in defining the different dialectal characteristics of Dari.

The other relevant issue is the existence of a Zoroastrian community in Yazd and Kerman. It is not exactly clear whether a Zoroastrian community already existed before the arrival of the immigrants to Yazd and Kerman. If this is the case, the newly arrived Zoroastrians in Yazd and Kerman would have adopted the regional language of the Zoroastrian community, but this language would have still been influenced by their original languages. If there was no Zoroastrian community before the arrival of the immigrants in Yazd and Kerman, they might have adopted the local language of the Muslim community. This language could have developed among the Zoroastrians in a new way and would have then been strongly influenced by the original languages of the new immigrants. If we assume the speakers of Zoroastrian Dari are immigrants from various regions of Iran, their language should reflect some characteristics of their original languages.

In order to explore this topic in more depth, different questions arise here: Does Zoroastrian Dari exhibit more Northwestern or Southwestern characteristics? According to the different origins of Zoroastrians of Yazd and Kerman, it can be claimed that Yazdi and Kermani represent two different language groups. For example, can it be concluded that Yazdi is more Northwestern than Kermani? If the original language of the immigrants of Yazd and Kerman is different, how did these dialects develop in two different ways? Assuming the origin of the Zoroastrians of Kerman is Khorasan, can we find similarities between Kermani Dari and the dialects spoken in today’s Khorasan province? Has there been any connection between Dari in greater Khorasan and Zoroastrian Dari, and if so, can the similarity in the names of these languages be found in their shared origin?

These are selected, relevant questions and for answering them, it is necessary to document all of the Dari sub-dialects. Let us look now at the main dialect and the sub-dialects of Zoroastrian Dari.


Zoroastrian Dari is one of the most unique Iranian languages on account of its large number of sub-dialects. As previously mentioned, there are two main dialects of Dari: Kermani and Yazdi. The dialect of Yazdi has many sub-dialects, while there appears to be only one dialect of Kermani. There were probably different dialects of Kermani in earlier periods of time, but today we find only one dialect. These possible former Kermani dialects outside of Kerman would be Jupāri, Qanātqestāni, and Esmāʿilābādi. The names Jupār, Qanātqestān, and Esmāʿilābād refer to three Zoroastrians villages near Māhān. The last Zoroastrian families left these villages, and most of them have moved to Kerman. There is no remnant of these sub-dialects anymore.

The Yazdi dialect of Dari has received much more scholarly attention than Kermani. The reason for this attention is that Yazdi is better preserved, and because of the existence of several sub-dialects, it offers more interesting areas of study for scholars. Yazdi Dari sub-dialects can be divided into four main groups: (1) High Dari “Malati,” spoken inside of Yazd; (2) dialects historically outside of Yazd, but currently spoken within the city Yazd or near Yazd; (3) Yazdi dialects spoken around Taft; (4) Yazdi dialects spoken around Ardakān and Meybod.

Malati. Many Zoroastrians believe that this dialect must be taught and standardized, and used as “high” Dari, because it is the dialect of the priests and their families. Malati is mostly treated as a high prestige dialect and is the language of both business and religion. As it possesses a large number of terms in every field, it can be used in various social and linguistic contexts. Malati is spoken in the following quarters of Yazd: Dasturān, Šesti, Sar-e lard-e āsiyāb (or Pošt-e ḵāna-ye ʿAli), Bozi, and Gav-e ḵar (or Gowdāl-e Šahriāri; cf. Mazdāpur, I, pp. 19-22).

Village dialects currently spoken in the city of Yazd. Dialects now located inside of Yazd are the following: K̠orramšāhi, Kuča Biyuki, Maryamābādi, Kasnaviyaʾi, Nersiābādi, Mehdiābādi, Aharestāni, Ḵeyrābādi, Qāsemābādi, Raḥmatābādi, and Ḥasanābādi. As for Moḥammadābād, Ābšāhi, Noṣratābād, and Dehno, the last Zoroastrian families have left these villages and their dialects cannot be found there anymore.

Dialects around Taft. The next group of Yazdi sub-dialects is spoken inside and around the city of Taft. This group consists of following sub-dialects: Mobārakaʾi, Zaynābādi, Čami, Ḵalilābādi, and Ḥosayni. Ḥosayni, is extinct.

Sub-dialects around Ardakān and Meybod. These are the villages around Ardakān and Meybod in which Zoroastrian families still live: Torkābād, Šarifābād, Mazraʿa Kalāntar, ʿAliābād, Elā(hā)bād, ʿAṣrābād, and Keštḵān-e Jaʿfarābād-e Šurak. In Aḥmadābād, Mehdiābād, and Šāhābād there are no longer any Zoroastrian families. Among this group, Šarifābādi and Elābādi are well known.

In the video to accompany this article (see https://youtu.be/KoTn9hk9VmE), the above-mentioned locations and sub-dialects are shown on a Google Earth map. For the symbols used in the video, see the legends below.

Map for Zoroastrian Dari Dialects



The next important factor is to assess the linguistic distance of Kermani from other Yazdi sub-dialects. To provide a better understanding of this topic, I mention some grammatical differences between Yazdi and Kermani. I refer to two Yazdi sub-dialects, Elābādi and Qāsemābādi, for the comparison of selected features.

Phonological differences. Historical ā is changed to ů in Yazdi. It seems that Kermani is closer to Persian here with this change (Yazdi důr-, dår-, present stem of důrtvun/důštvun “to have”; Kermani dār). This change is considered to be extremely trivial and takes place in many Iranian languages, maybe very recently.

Kermani ā in initial position can be compared to a bilabial spirant w, or to the vowel å in Yazdi (Kermani ārt “flour,” Yazdi vurt; Yazdi vudem “man, human,” Kermani ādem). In some Yazdi sub-dialects ä is found, whereas this vowel in Kermani changed to e (Yazdi čäm “eye,” Kermani čem; Yazdi ä “to, with, from,” Kermani e).

In the instances of enclitic pronouns, a final n in Kermani is attested, whereas there is no final n after the long vowel in Yazdi (see Table 4).

The infinitive ending in Yazdi is vun, whereas in Kermani mun is attested (Yazdi kartvun “do, make,” Kermani kartmun).

Initial h- is lost in Kermani, whereas it is preserved in Yazdi. This rule can be found in both Iranian words and Arabic loanwords (Yazdi ham- “together,” Kermani em-; Yazdi hämum “bath” (Ivanow, 1937, p. 68), Kermani emum; Yazdi heč, hečči “nothing” (Ivanow, 1937, p. 68), Kermani eči).

Morphological differences. Substantives. Substantives in Dari are very similar to those in Persian. In both of these languages, substantives have no distinction of grammatical gender. There are two numbers, singular and plural, and there is no morphological distinction between direct and oblique case. The prepositions and a postposition are used to mark the case of substantives. The plural suffix in Kermani Dari is , whereas in Yazdi, it is , -u, -o, or -un. The following Example 1 shows Kermani examples of singular and plural forms:


yāg “dish”

yāg-ā “dishes”


kar “lamb”

kar-ā “lambs”


The plural form of a few substantives ending with a vowel can also be constructed in a different way, with the suffix -gun.  Here are Kermani examples of singular and plural forms:


vača “child”

vača-gun “children”


In Kermani Dari, as in Persian, the postposition ro (Persian ) is used to mark accusative case, whereas in some Yazdi Dari dialects, the form e can be found instead of ro



bondā to-ro e-vin-a “I see you tomorrow.”




raven-e sar tevnen “heat oil!” (Mazdāpur, 2006, p. 20).


The dative case can be marked with the preposition e in Kermani (Persian be, Yazdi e, be “to”).



e me doru ma-vāj “don’t lie to me.”


In Yazdi Dari, in certain instances, other prepositions such as adu, xadu “with” (Persian ) are used instead of e or be “to.” 



ḥakim adu nukar=oš vut “the doctor told his servant” (Ivanow, 1935, p. 59).


The eżāfa is also expressed the same way as in Persian, with the addition of the suffix -i, -e (Persian -e) in both Kermani and Yazdi. 



ber-e mer “fire temple”


ḵor-i mas-i me “my grand sister”




bud-e ḵaš “a good smell” (Mazdāpur, 1995, p. 389).


Personal pronouns. There are some differences between personal pronouns in Kermani and in Yazdi Dari. The most important differences can be seen in the uses of the third person personal pronouns. In many Iranian languages, demonstrative pronouns are used as third person personal pronouns. In both Kermani and Yazdi, this characteristic can be found. In Kermani, the forms vin as third person singular and viyā as third person plural are attested. In Yazdi, the forms in, u “that” as the third person singular and iye “those” as third person plural are attested. In Table 5, the personal pronouns in Kermani are compared with those in Yazdi.

As mentioned earlier, in the case of enclitic pronouns in Kermani, the final n is attested, whereas in Yazdi, it has been omitted after the long vowel.

In some Iranian languages and dialects, for example, in Fars dialects, the pronominal enclitics become proclitics. It seems that there are also proclitics in Kermani. In Example 14 (below), proclitic om= is attached to the verb in Kermani, whereas in Example 15, =om as an enclitic is attached to the personal pronoun in Yazdi. The existence of proclitics is an important difference between Kermani and Yazdi. Proclitic pronominals are used in post-ergative patterns and are a combination of “and” and the pronominal clitics. Their existence can imply potential contact of Kermani Dari with Fars dialects. The following example illustrates the use of a proclitic in Kermani:


win oš=vo ve-ḵor-a “he wants to eat, he is going to eat”


Verb “to be,” verbal endings, verb conjugation. Table 5 shows the declination of both the independent form and suffixal form of the copula in Zoroastrian Dari.

It is interesting to observe that m, which very commonly occurs in the forms for the first person singular in Iranian languages, is not attested in Zoroastrian Dari. In this regard, Dari is similar to Nāʾini. The forms for the third person singular and plural are different from the forms in Persian. The existence of -h- in the independent forms can also be found in Middle Persian verbal forms. The forms Yazdi -un, -uné, -oné, -ůné, and Kermani -on usually occur after a consonant, while -na, -ne occur after a vowel. In terms of the verb “to be,” there are only a few phonological differences between Yazdi and Kermani, especially in the instance of the third singular enclitic form. The forms of neʾim, neʾi and neʾen are found only in Yazdi.

Ergativity. As for past transitive constructions, Dari shares a parallel development with the Central group of Kurdish, especially Sorani. Since Dari has lost case distinctions similarly to Sorani, the past transitive construction has developed in such a way that pronominal enclitics are used to mark the agent and no longer indicate a particular case distinction. The agreement of enclitic pronouns with the agent in Kurdish has already observed by T. Bynon (p. 217) and T. Jügel (p. 147). In his study, G. Haig (pp. 288-89) speaks about “cross-reference” between the agent and the enclitic pronouns in the Solaymāni dialect of Sorani. D. N. MacKenzie (1961a, pp. 107-8) called enclitic pronouns the agent markers in the past tense of transitive verbs. It appears that in Dari, every single past transitive construction requires the presence of an enclitic pronoun that refers to the clause agent, regardless of whether the agent is lexically realized elsewhere in the clause or not. In what follows, I compare Dari and Sorani Kurdish to illustrate this similarity (for a list of the abbreviations used in the following examples, see below, Bibliography):










“You worked” (Yügel, p. 147).


Yazdi Dari:








“They narrated” (Ivanow, 1937, p. 68).


In Example 12, the agent is not expressed as an independent lexical form, while there are other examples of the existence in the same clause of both an independent lexical form expressing the subject and enclitic personal pronoun as a subject marker (see Example 13). In both Ex. 11 and 12, the agent is indexed by an enclitic pronoun (CP), which is a form always used in oblique function. It must be added that, in a language like Sorani, besides the abovementioned enclitic pronouns, there are also personal endings. This type of personal marker is missing in Dari. For the function of personal endings as a personal subject marker in Sorani, see Jügel (pp. 152-56).

In Example 13, the agent is not expressed, whereas in Example 14, the agent, that is, märdumuni por “a group of people” is expressed, and the enclitic pronoun šů also refers to it.






























“a crowd of people took school children with them” (cf. Ivanow, 1937, p. 68).


In Kermani, the patient is marked with the object marker ro (Persian ), which clearly shows influence from Persian. In Yazdi, the object marker is not used. The following examples illustrate the past transitive construction in Yazdi, Kermani, and Persian.











“I saw you.”


Qāsemābādi and Elābādi:










“I saw you.”












“I saw you.”


Lexical differences. In terms of the lexicon, both dialects show some differences, which can even result in a degree of mutual incomprehensibility between speakers. The following are a few examples of these lexical differences: Yazdi herdů, Kermani bondā “tomorrow”; Yazdi aḵova, Kermani govāf “yawn” (Sorušiān, 1978, p. 4); Yazdi nůḵaš, Kermani vezir “sick”; Yazdi menagen Kermani ānomid “good, fine” (Sorušiyān, 1978, p. 156); Yazdi hůsut, vůsut, Kermani ḵāteri “ashes.”

In the case of kin relationship terms, there are a few differences between Kermani and Yazdi. In Table 6, a list of terms is found. It seems that Kermani pōrer and doter are generalized from other family terms, as in Persian.

Various phonological, morphological, and syntactic differences between Yazdi and Kermani clearly support the existence of two independent dialects. It seems that the Yazdi dialect of Dari is more conservative than Kermani in many respects because of the special conditions of the Zoroastrian community in Yazd and the surrounding areas.

What I have shown here are various differences between Kermani and only two Yazdi sub-dialects. It is relevant to examine similar characteristics in other Yazdi sub-dialects. It would be particularly relevant to document Malati, which is regarded as the high form of Dari and the language of priests. Šarifābādi is another relevant sub-dialect, because it possesses many archaic characteristics.


An important topic for linguistics of Iranian languages is the family relationships of these languages and the model illustrating the kinds of relationships. Traditionally, the relationship between languages is shown by means of a family tree, which has a Western and an Eastern Iranian branch. Each of these in turn is subdivided into a Northern and a Southern branch. This model is offered on the basis of the analysis of certain Middle Iranian text materials, which were found around 1900 (see TURFAN EXPEDITIONS; DUNHUANG i. The cave sites; Manichean texts; SALEMANN, CARL HERMANN). So far as Western Iranian is concerned, these texts are in two different Western Iranian dialects (today known as Manichean Middle Persian and Parthian), which differ in a number of characteristics (isoglosses). These isoglosses are used to identify possible subgroups and to determine the position of a particular language in relationship to other languages. This tradition, which was first introduced by P. Tedesco and others for the purpose of distinguishing the languages present in the newly discovered texts, does not seem to be a very successful method for determining the dialectology of Iranian languages, because the features by which Middle Persian and Parthian agree do not figure in this list of isoglosses. There are also other problems in this model. After new findings, especially those of Bactrian materials, it became clear that a new method should now be offered to classify the Iranian languages and to illustrate a model for their relationship. For example, Bactrian belongs to the Eastern branch, but it also possesses many characteristics of the Western group. This shows that this language can have had contact with Western languages in a period of its existence. Many examples from different languages prove that languages encounter others during particular periods of history and in particular places, and thus influence each other. To clarify the family relationships of the Iranian languages, every language is important and must therefore be taken into account. This has been done in the case of many languages, but it is now necessary to study the languages which have not yet been sufficiently researched. One of these languages is Dari, which has particular importance for this purpose. There are many unclear issues regarding the classification of this language. The position of this language in relationship to other Iranian languages is still unknown to a large extent. Dari has generally been considered to be closely related to Persian or to Central dialects in the earlier works. This section aims to present some important features of Dari that have not been dealt with in previous works. As a step with this purpose, it is relevant to look at Western Iranian isoglosses.

L. Paul (1998) presented Western Iranian isoglosses in his article. J. Gippert et al. (2007-08) reviewed Paul’s article and discussed the dialectal position of Zazaki (see Dim(i)li). A. Korn (2003) examined the position of Baluchi among Western Iranian languages and presented Western Iranian isoglosses in chronological order. She indicated that, in the majority of the isoglosses that are traditionally used to distinguish Northwest from Southwest, the Old and Middle Persian outcomes of certain proto-Iranian sounds (or sound combinations) can be interpreted as innovations, whereas Avestan and Parthian preserve the proto-Iranian state of affairs. She added that the Northwestern Iranian languages show considerably fewer innovations and preserve the developments that can be found in Parthian. G. Windfuhr (1992, 2009), P. Lecoq (1989), D. Stilo (2007), and H. Borjian (2011) offered different isoglosses of Central group dialects and subgroups.

For the phonological similarities and differences of Dari with NW and SW Iranian languages, we first look at the following list, which presents the Western Iranian isoglosses in Old, Middle, and New Iranian, with related examples.

According to the Western Iranian isoglosses in the Old, Middle, and New Iranian periods, it seems that Dari is closer to the NW than to the SW group, in regard to the development of PIr. (D-Y ruvås, D-K rubā, Parth. rōbās, Bal. rōpāsk, NP rōbāh “fox”), (D-Y, D-K zun-, Av. zanā-, Z zan-, Bal.zān, Kr. zān-, NP dān, MP dān, OP dānā- “know”), *śṷ (D-Y sva, D-K saba, P ispag, South Z espe, NP sag “dog”), *θr (D-Y pōr, D-K pōrer, P puhr, Bal. pus(s)ag (loanword), Kr. Abduʿi pus (loanword), NP pesar, MP pus “son”; D-Y, D-K se, P hrē, Z hire, NP and MP “three”), / V_V (D-Y, D-K ruj, Z rôž, rôj, Bal. rōč, NP rōz “day”), *ṷ- (D-Y, D-K vin, Z vēn-, vin-, Bal. gind-, Kr. bîn; NP bin “see”) and *dṷ (D-Y, D-K bar, P bar, Central Z ber, NP and MP dar “door”) and shares more similarity with Zazaki from NW group.

In addition to the abovementioned isoglosses, regarding Old Iranian d, Dari also shares similarities with the NW group. Old Iranian d can be found in Kermani and few Yazdi Dari, for example, in kod “when” (P kaδ, Bal.kadi, NP key, MP kay). In two Yazdi sub-dialects, both đ and z are attested (Elābādi kođ, Qāsemābādi kāz).

In the case of the development of *i̯- to j- (D-Y jedů, D-K jodā (probably loanword)  P yuδ, Z jiyā, NP and MP jud(ā)g “separate”; Isfahani Jewish yuš/ā, D yus- “boil”; Stilo, p. 9), Dari shows the same result as Persian.

Verb forms have often been used to demonstrate the dialectal positions of Iranian languages. For example, the present stem of the verb “to do” can be used as an isogloss for the NW group (see Tedesco, p. 223). Another example would be the verb “to come” (see MacKenzie, 1961a, pp. 74-75). Paul (2003, pp. 62-63) considered these two verbs together with four other verbs: “to have, keep,” “to say,” “to go,” “to fall” to demonstrate the dialectal position of Baluchi. Stilo (2007, p. 9) mentioned that the two verbs “buy” and “cut” retain a final -n in their present stems in the NW group. The present stem of the verb “say,” that is, vaj- in Dari, is close to the NW group (P vāž, Z vāž), whereas in the SW group *gaub- (NP ) is attested.

The past stem of the verb “to fall” is kapt in Dari, and this form is well attested in the NW group. In contrast, one finds that in the SW group, the forms are usually derived from *ava-pat- (NP oftād).

In the case of the two verbs, “buy” hrin- and “cut” brin-, Dari here apparently preserves the old sequence of rn and does not show the specifically Persian sound change of rr.

In Dari, another development of Old Iranian kṛnu- can be found in comparison to Persian. In Persian, the present stem of the verb kardan “to do, to make” is kon-, while in Dari, kar- is attested.

In a similar way in which they provide isoglosses to distinguish the NW and SW groups, some scholars also provide isoglosses for Central dialects. In spite of all the isoglosses that have been offered for the Central dialects, there is still no clear distinction between NW and Central Dialects. The internal grouping of the Central Dialects displays considerable differences and the lack of isoglottic bundling. As the majority of Central dialects have not yet been sufficiently studied, it is not easy to make a more exact judgment regarding isoglosses of this group. Windfuhr (1992) and Stilo (2007) have discussed some of these isoglosses. I focus on these characteristics in order to show the similarities and differences of Dari in relation to the Central dialects.

In the case of the prefix *fra-, the most common development is r in the Central group. Both Kermani and Yazdi Dari exhibit a different development to hr. Another isogloss is the general marker of the perfective tenses, including the perfective subjunctive ba-/be-. It is used in the majority of Central group languages. In both Kermani and Yazdi Dari, the prefix be- has undergone change to ve, and it seems to have a similar function to the prefix be in Persian. An example in Dari is ve-pars-e “he/she would ask.”

According to Windfuhr (1992), another isogloss is the formation of the durative present and past. There are two distinct patterns, as the prefix *at- (also attested in different Central Dialects as the forms ed, et, e, a) or as the enclitic particle -e. The function and characteristics of this suffix and affix are discussed in detail in Stilo (2007, pp. 13-15). According to Stilo, in the Isfahan area, the suffixed form -e is used to mark durative, whereas in other areas the prefix e(t)- is found. There are two patterns, depending on whether or not the verb is negated. On verbs without negation, the durative marker in both Yazdi and Kermani is a prefixed et- before a vowel and e- before a consonant. However, if the verb is negated, the durative marker appears as a suffix. The following examples illustrate the prefixed and suffixed forms: Yazdi Dari: et-ů-t, et-ů, Kermani Dari et-ā, et-ā-d “he/she comes” but Yazdi Dari n-ů-t-e, Kermani Dari n-ā-t-e “he does not come.” Regarding the pronominal form referring to “we,” both Dari and Nāʾini exhibit a typical SW form. In the case of the demonstrative pronoun, “this,” Dari and K̠uri show similarities. In both Dari and K̠uri, a m- stem demonstrative can be found: Yazdi Dari min, mo; Kermani Dari min, , K̠uri em. Dari also appears to be similar to Āštiāni-Tafreši, Isfahani, K̠uri, and Sivandi, in that Dari also has a b- stem expressing “become.” With the form ůsu “now,” Yazdi Dari appears to be more similar to Isfahani and Āštiāni-Tafreši. Regarding the adjective for “big,” Dari appears to be closer to Nāʾini and Maḥallāti to K̠vānsāri groups (see also Stilo, Table 2). In all these languages, mas “big, great” is attested. In the case of the verbal ending for the first person singular, both Yazdi and Kermani Dari and Nāʾini present an innovation: Yazdi Dari -e, -a, Kermani Dari -e, -a, and Nāʾini -i(e). In all other NW and Central dialects, the forms   ‑(u)m, -(i)m, -(u)n, -(o)n are attested.


Work on a truly comprehensive documentation of Zoroastrian Dari requires various types of expertise in a multitude of disciplines. Basic linguistic expertise in transcribing or transliterating is not sufficient. In addition to these requirements, the knowledge of anthropology and of historical and religious studies is required. The special character of Zoroastrian society and the status of Zoroastrianism as a religious minority create a challenging and complicated situation for fieldwork. Some of the theoretical and practical problems and factors have already been mentioned in the present article. There are also certain ethnic problems and limitations, which are related to the circumstances of the Zoroastrian community in Kerman and Yazd. In the present article, I have tried to present the importance of selected disciplines for documenting and analyzing Zoroastrian Dari. Such disciplines include the history of the Zoroastrian community and cultural and social anthropology. In addition to these factors, selected grammatical and dialectal characteristics of Zoroastrian Dari have been mentioned. In this regard, it should be emphasized that the historical and cultural background of Dari can influence the structure of the language, its sub-dialects, and its dialectology.

It initially appears that Dari exhibits many characteristics of the NW Iranian group. Among the Central dialects, Dari appears to share a high number of isoglosses with Āštiāni-Tafreši, although with regard to geographic distance, Dari is spoken in an area closer to Nāʾin. With regard to phonology, morphology, and syntax, it appears that Dari shares features with languages of the NW group, especially with Zazaki and Sorani.

Sharing of both NW and SW features draws our attention to the fact that perhaps the immigrants to Yazd and Kerman came from different provinces of Iran. A question is raised here: can we claim that one of these dialects has more NW or more SW features? To clarify this question, I present an example that was mentioned earlier under examples for isoglosses of Western Iranian languages: Yazdi Dari sebul and Kermani Dari svarz “spleen.” In this example, Yazdi Dari shows a typical Southwestern Iranian development, whereas the Kermani form shows a Northwestern development. We know that seporz is attested in Persian, which is a loanword from a Northwestern language and cannot be a source for Yazdi sebul. Could this example show that Yazdi Dari has been influenced to a greater degree by the SW group than Kermani has been? Surely one example is not sufficient to make a decision, and I will attempt to find more such characteristics and discuss them in my project. There are relevant questions of this type that can be answered after documenting all of the Zoroastrian Dari sub-dialects.

Informants. Informants for Kermani: Farangis K̠adivi, 83 years old, born in Kerman, lives in Kerman, was a teacher; Zemorrod Sohrābiān, 80 years old, born in Kerman, lives in Kerman, was a teacher; Purān Javānmardiān, 82 years old, born in Kerman, lives in Tehrān, is a housewife.  Informants of Yazdi: Qāsemābādi Yazdi: Sohrāb Zand-K̠āvari, 49 years old, born in Qāsemābād, lives in Kerman, is an electrical engineer; Elābādi Yazdi: Mahvaš Bāstāni, born in Elābād, lives in Kerman, is a housewife; Mr. Bāstāni, born in Elābād, lives in Yazd, is a director of the Purčistā rest home in Yazd.  Informant for Zazaki: Mesut Keskin.  Informant for Sorani and Hawrāmi (see Avromani): Diako Nāhid.

For a sample of Zoroastrian Dari dialect in Kerman, you may view the following video:




enclitic pronoun


New Persian




Object Marker














Middle Persian









H. Borjian, “Kashan ix. The Median Dialects of Kashan,” in EIr. XVI/1, pp. 38-48; at http://www.iranicaonline.org/articles/kashan-ix-the-median-dialects-of-kashan.

C. Bosworth, The Medieval History of Iran, Afghanistan, and Central Asia, Cambridge, 1977.

T. Bynon, “The Ergative Construction in Kurdish,” BSOAS 42, 1979, pp. 211-24.

J. K. Choksy, Conflict and Cooperation. Zoroastrian Subalterns and Muslim Elites in Medieval Iranian Society, New York, 1997.

E. B. N. Dhabhar, The Persian Rivayats of Hormazyar Framarz and others. Their Version with Introduction and Notes, Bombay, 1932.

F. Firuzbaḵš, Barrasi-ye sāḵtemān-e dasturi-e guyeš-e Behdinān-e šahr-e Yazd (The study of grammatical structure of the Behdinān dialect in Yazd), Tehran, 2000.

W. Geiger and E. Kuhn, eds., Grundriss der iranischen Philologie, 2 vols., Strassburg, 1895-1904.

T. E. Gindin, “Yazd iv. The Jewish Dialect of Yazd,” EIr., online edition, 2009, at http://www.iranicaonline.org/articles/yazd-iv-the-jewish-dialect-of-yazd.

J. Gippert, N. Himmelmann, and U. Mosel, Essentials of Language Documentation, Berlin, 2006.

Idem, “Zur dialektalen Stellung des Zazaki,” Die Sprache 47, 2007-08 1, pp. 77-107.

G. Haig, Alignment Change in Iranian Languages, Berlin, 2008.

J. H. Hill, “The ethnography of language and language documentation” in J. Gippert et al., eds., Essentials of Language Documentation, Berlin, 2006, pp. 113-28.

M. Ḥasandust, Farhang-e taṭbiqi-możuʿi-ye zabānhā wa guyešhā-ye irāni-ye now I, Tehran, 2011.

A. Houtum-Schindler, “Die Parsen in Persien, ihre Sprache und einige ihrer Gebräuche,” ZDMG 36, 1882, pp. 54-88.

W. A. Ivanow, “The Gabri dialect spoken by the Zoroastrians of Persia,” Revista degli Studi orientali 16, 1935, pp. 1-17, 18-6; 17, 1937, pp. 68-106; 18, 1939, pp. 107-64.

T. Jügel, “Ergative remnants in Sorāni Kurdish?” Orientalia Suecana 58, 2009, pp. 142-58.

J. Kestenberg Amighi, The Zoroastrians of Iran: Conversion, Assimilation, or Persistence, New York, 1990.

A. Korn, “Balochi and the Concept of North-Western Iranian,” in Agnes Korn and Carina Jahani, eds., The Baloch and Their Neighbours: Ethnic and Linguistic Contact in Balochistan in Historical and Modern Times, 2003, Wiesbaden, pp. 49-60.

Idem, Towards a Historical Grammar of Balochi: Studies in Balochi Historical Phonology and Vocabulary, Wiesbaden, 2005.

K. J. Krahnke, “Linguistic Relationships in Central Iran,” Ph.D. dissertation, University of Michigan, 1976.

P. Lecoq, “Les dialectes du centre de l’Iran,” Compendium Linguarum Iranicarum, 1989, pp. 313-26.

D. L. R. Lorimer, “Notes on the Gabri Dialect of Modern Persia,” Journal of the Royal Asiatic Society of Great Britain and Northern Ireland, 1916, pp. 423-89.

Idem, “Is There a Gabri Dialect of Modern Persia?” Journal of the Royal Asiatic Society of Great Britain and Northern Ireland, 1928, pp. 287-319.

D. N. MacKenzie, Kurdish Dialect Studies I, London, 1961a.

Idem, “The Origins of Kurdish,” Transactions of the Philological Society, 1961b, pp. 68-86.

K. Mazdāpur, Vāža-nāma-ye guyeš-e behdinān-e šahr-e Yazd, 2 vols., Tehran, 1995-2006.

F. F. Omar, Kurdisch-Deutsches Wörterbuch (Nordkurdisch/Kurmancî), Berlin, 1992.

M. Ošidari, Farhang-e vāžahā wa zamān-vāžahā-ye guyeš-e Zartoštiān-e Kermān, Tehran, 2008.

L. Paul, “The Position of Zazaki among West Iranian Languages,” Proceedings of the Third European Conference of Iranian Studies I, 1998, pp. 163-77.

Idem, “The Verbal System,” in Agnes Korn and Carina Jahani, eds., The Baloch and Their Neighbours: Ethnic and Linguistic Contact in Balochistan in Historical and Modern Times, 2003, pp. 61-71.

R. Šahmardān, Tāriḵ-e zartoštiān, farzānegān-e zartošti, Tehran, 1984.

N. Sims-Williams, “Notes on Manichaean Middle Persian Morphology,” Studia Iranica 10, 1981, pp. 165-76.

P. O. Skjærvø, “Case in Inscriptional Middle Persian, Inscriptional Parthian and the Pahlavi Psalter,” Studia Iranica 12, 1983, pp. 47-62, 151-81.

J. S. Sorušiān, Farhang-e behdinān, Tehran, 1978.

Idem, Tāriḵ-e zartoštiān-e Kerman dar in čand sada I, Kerman, 1991.

M. Stausberg, “From Power to Powerlessness: Zoroastrianism in Iranian History,” in A. N. Longva and A. S. Roald, eds., Religious Minorities in the Middle East: Domination, Self-Empowerment, Accommodation, Leiden, 2012, pp. 171-93.

D. Stilo, “Isfahan xxi. Provincial Dialects,” in EIr. XIV/1, 2007, pp. 93-112, at http://www.iranicaonline.org/articles/isfahan-xxi-provincial-dialects.

Tāriḵ-e Sistān, ed. M.-T. Bahār, Tehran, 1935; English tr. Milton Gold, Rome, 1976.

P. Tedesco, “Dialektologie der mitteliranischen Turfantexte,” Monde Oriental 15, 1921, pp. 184-258.

E. M. R. Unvala, Dārāb Hormazyār’s Rivāyat II, Bombay, 1922.

G. Windfuhr, “Behdinān Dialect i. Lan,” in EIr. IV/1, 1989, pp. 105-8; at http://www.iranicaonline.org/articles/behdinan-dialect.

Idem, “Central Dialects,” EIr. V/3, 1991, pp. 242-52; at http://www.iranicaonline.org/articles/central-dialects.

Idem, “Dialectology and Topics,” in G. Windfuhr, ed., The Iranian Languages, 2009, pp. 6-42. 

(Saloumeh Gholami)

Originally Published: January 27, 2016

Last Updated: January 27, 2016

Cite this entry:

Saloumeh Gholami, “ZOROASTRIANS OF IRAN vi. Linguistic Documentation,” Encyclopædia Iranica, online edition, 2016, available at http://www.iranicaonline.org/articles/zoroastrians-in-iran-06 (accessed on 27 January 2016).