This article examines the linguistic features of Khorasani Persian as spoken and written in the vast region stretching from Qumes to Marv, which in the inscription of Darius (q.v.; see also BISOTUN iii) was called Parθava, or “Parthia.”
The adjectival forms of Pahlaw in Middle Iranian languages are Pahlawānīg and Pahlawīg. The more recent forms of the word Pahlaw in Persian and Arabic are Pahla and Fahla, and their adjectival forms Pahlavi and Fahlavi (see FAHLAVĪYĀT). Though Persian Pahlav and Arabic Fahla are the formal continuations of Old Persian Parθava and Middle Persian Pahlaw, they no longer refer to the historical region of Parthia. According to Ebn al-Moqaffaʿ (q.v.; d. 139/757) as quoted in the Fehrest (q.v.; Ebn al-Nadim, ed. Tajaddod, p. 15, tr. Dodge, I, p. 24), Fahla consisted of the regions of Isfahan, Ray, Hamadān, Māh Nehāvand, and Azarbaijan, that is a region comprising Media (q.v.).
The word Pahlawānīg is attested in a Manichean Middle Persian text (Andreas and Henning, pp. 302-3): When the Messenger of Light (= Māni [q.v.]) was in Holwān, he called Mar Ammō, one of his companions who knew the Pahlawānīg language (= Parthian), and sent him to Abaršahr (q.v.), i.e, the Nishapur province in western Khorasan. The Sogdian version of the same text also mentions the Pahlawānīg language (Sogd. pγlʾwʾnʾk), a valuable testimony on the linguistic situation of Iran in the third century CE (Henning, 1958, p. 94; Lazard, 1971, p. 364).
Manichean Middle Persian embodies Middle Persian in its original, provincial purity; the Middle Persian of the books, as the common language of the Sasanian Empire and the language of education of its priests and court singers, was exposed over time to the influence of the older Parthian vernacular language. It eventually became a dialect that lost the peculiarities of its southwestern Iranian origin (Sundermann, 1989, p. 139).
During the Sasanian period (224-650 CE), Pārsīg (usually called Middle Persian or, erroneously, Pahlavi), that is, the language attributed to the region of Pārs, was the official language of Iran. Both Pahlawīg and Pārsīg are connected to the western branch of the Iranian languages (Pahlawīg to the northern branch and Pārsīg to the southern) and are two dialects of the same language. During most of the Arsacid period (250 BCE-226 CE), Pahlawīg, along with Greek, was the official language. Here, we will use Parthian instead of Pahlawīg, and Middle Persian for Pārsīg.
The Manicheans wrote their propagandistic pamphlets intended for Pahla in Parthian. However, when Middle Persian spread and obtained official status, the use of Parthian gradually diminished, so that, apparently by the 6th century, nothing seems to have been written in Parthian as a living language. A few short private inscriptions in Parthian language and script were found on rock-faces in southern Khorasan, that is, within the territory of Parthia proper; but it is thought that none is later than fourth century CE. Antoine Ghilain (q.v.) believed that the disappearance of Parthian as a living language took place in the 5th century (p. 28). Moḥammad b. Aḥmad Ḵᵛārazmi (d. ca. 380/990), however, writes (p. 117) that the Sasanian kings conversed in this language in their gatherings (majāles). If this is true, then apparently, because of its prestige, Parthian was still used in court circles as an esteemed language for some time after having lost popular currency (see Lazard, 1971, pp. 378-80).
In the Islamic period, we see in Khorasan the use of a developed form of Pārsīg, or Pārsi, which would later be called Fārsi, or Pārsi/Fārsi-e Dari, and in Arabic al-fāresiya al-dariya (see DARI). Ebn al-Moqaffaʿ, an Iranian who wrote in Arabic, said that Pārsi-e dari was the language in Sasanian times of the capital city Madāʾen/Ctesiphon (q.v.); it was also one of the languages of the people of Khorasan and the east, and was based on the language of Balḵ (q.v; apud Ebn al-Nadim, ed. Tajaddod, p. 15, tr. Dodge, I, p. 24). Wāʿeẓ Balḵi (fl. 1214) also stated, quoting Nażr b. Šomayl (122-204/740-820), that Dari is the language of the people of Balḵ (pp. 29-30). Ebn al-Moqaffaʿ’s considering Dari to have been dominated by the language of Balḵ apparently means that the standard form of Dari was the one current in Balḵ. Abu ʿAbd-Allāh Maqdesi’s statement (p. 334) that the language of Balḵ is “the best” seems to support this view. Ebn al-Moqaffaʿ’s description of the languages of Iran—Pahlavi, Pārsi, Dari, Soryāni, and Khuzi (Ebn al-Nadim, ed. Tajaddod, p.15, tr. Dodge, I, p. 24) indicates the variety of languages existing in Iran at the end of the Sasanian period.
Ebn al-Moqaffaʿ’s reference to Balḵ—and, later, that by Nażr b. Šomayl—shows that at the end of the Sasanian period Dari had spread from Ctesiphon (q.v.), the Sasanian capital, to Khorasan. Joseph Markwart (q.v.) states that, around the end of this period, Middle Persian had probably reached Ṭoḵarestān (the region of Balḵ; see BACTRIA), and that later, apparently at the beginning of the Arab domination, appeared as the language of general exchange and everyday interactions (p. 89). Arthur Christensen (q.v.; I, p. 5) regarded the transfer of troops to the east and the establishment of military bases there for the purpose of resisting invaders from Central Asia as the predominant factor leading to the general use of Dari in Khorasan.
We know that attacks on Khorasan by the Chionites (q.v.) began in the 4th century. The first entrance of Dari into Khorasan probably took place in the same century or, as is more likely, beginning in the 5th century. In Khorasan, Dari was profoundly influenced by Parthian. In addition to a great number of words, it was also influenced by Parthian with respect to phonetics and phonology as well as grammar. Wolfgang Lenz (q.v.) listed many Parthian words used in Ferdowsi’s Šāh-nāma (pp. 251-316). Of course, a number of Parthian forms and roots are also notable in Middle Persian—or, in the terminology of some scholars, “Zoroastrian Pahlavi” (=Middle Persian) or “book Pahlavi” (Pahlavi-ye ketābi). These elements probably entered the official language from the very beginning of Sasanian rule, when Pārsīg became the official language of the court. For example, the word čiš, meaning čiz ‘thing’, entered Middle Persian from Parthian; the original southern form, tis, also appears in some Judeo-Persian texts, and it is its more ancient form, *tsis, which is today pronounced t s es in Davān, a village to the northeast of Kāzerun (q.v.). The word čiz is probably itself also an altered form of Parthian čiš. The aorist root of the verb dādan ‘to give’ is day in Manichean Middle Persian, but dah in Parthian. The Persian word ranj ‘trouble, suffering’ is also ranj in Parthian, and ranǰ in Middle Persian, taken from Parthian; the Manichean Middle Persian form is ranz. The word panj ‘five’ is panj in both Parthian and Persian, but panz in Manichean Middle Persian. The forms dah, ranj, and panj did not therefore enter New Persian directly from Parthian, but through Middle Persian.
We do not have much information about Dari in the late Sasanian period. The oldest examples of Dari Persian are phrases attributed to Sasanian rulers, or others, found in Arabic sources (see Ṣādeqi, 1978). Apart from these phrases, we also have two letters in Judeo-Persian, in Hebrew script. One of these letters is deficient, as four sides of it have perished; therefore many sentences are incomplete. This letter was found by Sir Aurel Stein (q.v.) in the late 19th century in the ruins of Dandān Öilïq (q.v.), and it is thus known as the “Dandān Öilïq Letter.” A number of scholars have studied this letter, including Bo Utas, to whose transcription and translation of the text Gilbert Lazard (1988) has provided some emendations. The language of this letter is Dari Persian, but very ancient, perhaps from the mid-8th century. The ergative constructions (q.v.) of transitive verbs in the past are absent from this letter. The past and past perfect tenses correspond to their forms in classical Dari Persian: foruḵta bud ‘he had sold’; foruḵta buda ast ‘he had had sold’. But instead of ke at the beginning of a relative clause, as in Middle Persian, ī is used: kār-ī farmudi-aš saḵt konom tā karda bovad ‘I will insist till the work you ordered is done’. This construction is also seen sometimes in texts from the 11th and 12th centuries (see below). There is only one ancient word in this letter that is not seen in classical texts: bendom ‘I find’, bendādom ‘I found’ (MPers. windādan). W. B. Henning (q.v.) thought that the 8th century was too early for the writing of this letter (Henning, 1958, pp. 79-80); however, the extreme scarcity of Arabic words in it may support the suggestion of the 8th century as the date of its composition. The word rikēbayn, a mispronunciation (emāla: pronouncing ā as ē) of rekābayn ‘stirrups’ is apparently the only Arabic word. On the other hand, there are two Sogdian words, cmkwy ‘harp’ and ʾndryk ‘eunuch,’ in it, which may indicate the letter’s connection with some decades later than the beginning of the 8th century.
Several years ago, the first page of another commercial letter in Judeo-Persian was found, which is nearly complete and intact. Two Chinese scholars, Shi Guang and Zhang Zhan, published this letter in 2008 and added detailed comments on it in Chinese (see also Yoshida). In Lazard’s 2014 article, based on the reading of “an unknown individual” (Lazard’s words), brief information was given on the letter’s linguistic characteristics, which are summarized here: They include early pronunciations, such as abā ‘with’, andar ‘in’, and ayāftan ‘to find’; the conditional mode in Middle Persian forms such as agar…rasād ‘should he/it…arrive’, and nāmada bād ‘let it not happen’; passive adjectives are used in the present perfect without the suffix –a: nebešt budi ‘it was written’; the sign for the enclitic eżāfa (q.v.) is -ǐ instead of ke at the beginning of a relative clause, such as: ān-e nebešt budi ‘that which was written’ (Lazard, 2014, p. 90). Kasra instead of ke is also seen in some Persian texts (see Ṣādeqi, 2016, pp. 3-7). Given the fact that a closed community, such as the Jewish community in a predominantly Muslim society, would tend to be more conservative with respect to language, one cannot date such documents with any degree of precision. What is certain is that this letter has an early date, and it was probably written around the 8th century.
Another early text, preserved with the Manichaen fragments from Turfan in Berlin, is in a non-Persian script. It consists of two pages of a translation of the Psalms into Persian in Syriac script. In 1915, Friedrich Müller published the upper section of these two pages (pp. 215 ff.). Much later, Werner Sundermann discovered another part of these two leaves, and published all the existing segments with a re-examination of the text published by Müller (1974, pp. 441-52). In 2011, Nicholas Sims-Williams undertook a new transcription, translation, and publication of this text. It, too, has early characteristics, such as the use of the gerund without the suffix -a, as in payrāst budand ‘they were prepared’; the use of preposition p=pe or pa instead of be, which is seen in some other Persian texts such as the Tafsir-e Qorʾān-e pāk and the Judeo-Persian text published by David Neil MacKenzie (q.v.; 1968); the use of the present stem of the verb dādan with t instead of d, which is still used in Tajikistan, Afghanistan, and eastern Iranian Khorasan: thyḏ ‘give’, thḏ ‘gives’, etc.; the use of some words that apparently are not used in Dari Persian, such as pāḏyāvand ‘strong’, which occurs also in MacKenzie’s text: ram meaning mardom ‘people’; gorāḡ for kalāḡ ‘crow’ (seen also in Baluchi); mānešn for ḵāna ‘house’ (evidently a historical spelling for mānešt); jud-ābitar for ḵošktar ‘drier’; kāmaḏ, from the verb kāmestan meaning ḵᵛāstan ‘to want’; ē-rā-ke for zirāke ‘because’. Only two Arabic words are found in this text: ḥadd ‘border’ and jomla in jomlagi ‘summation’. This translation is probably from the 9th century; but in it some dialect words may also be seen.
A number of texts from the 10th and 11th centuries also contain early and dialectal characteristics, some of which will be dealt with at the end of this article. But many Khorasani texts of this period are written in standard classical Persian; and with the exception of early words, which later fell completely out of currency, there are no dialect words in these texts. Various classical, standard, 10th and 11th-century texts substantiate this claim: Moḥammad Ḡazāli (q.v.; 450-505/1058-1111) wrote two books, Kimiā-ye saʿādat (q.v.) and Naṣiḥat al-moluk, in standard classical Persian, in which almost no dialectal characteristics can be seen, or else such words are very few. For instance, such words in the Kimiā, which became virtually obsolete after the 12th century, include: āzmāneš, for āzmāyeš ‘trial, test’; asta for hasta ‘fruit stone, pit’; bestāḵ along with gostāḵ ‘overly familiar (persons)’, ‘bold, rude’; bonješk for gonješk ‘sparrow’ (Naṣiḥat al-moluk, p. 184); neḡuša, meaning “secretly listening to someone, eavesdropping,” for niōša; bāzidan for bāḵtan ‘playing; gambling’ (in the context of chess, gambling and pigeon-flying), from Middle Persian wāzidan; and some infinitives and analogical past tense verbs such as sarāyidan for sorudan ‘to sing, recite’ and jahidan for jastan ‘to jump’. The Middle Persian consonant w changes to g or b in New Persian; and the forms bestāḵ and bonješk cannot be considered dialectal.
In the Naṣiḥat al-moluk as well there are only a few early and dialectal forms, such as kuhan=kwhn for kohan ‘old, ancient’ (p. 67); kahriz for kāriz (q.v.) ‘subterranean canal’ (pp. 189, 242); āḡāz for āvāz ‘voice, sound’ (p. 297); šugin for šuḵgen ‘dirty’; dig-ruz for diruz ‘yesterday’; zafān/zufān for zabān ‘tongue, language’ (p. 169).
The situation is the same for the language of Neẓām-al-Molk Ṭusi’s (408-85/1018-92) Siar al-moluk (qq.v.). It has only a few early or dialect words, such as: rawišn for raveš ‘conduct’; bērān for virān ‘ruined’; ādēn for āyin ‘custom’; tābān for tāvān ‘compensation’; barzidan for varzidan ‘work, practice’; avām for w/vām ‘loan, debt’, from Middle Persian abām; and F/Vistā for Avesta (pp. 258, 265).
It is also notable that Ferdowsi’s Šāh-nāma, composed earlier than Siar al-moluk, was also written in standard classical Persian. Some particular forms, such as miža for moža ‘eyelash/es’; goyāzanda for godāzanda ‘boiling; burning’; bad-Irān for ba-Irān ‘to Iran’; bad-Afrāsiāb for ba-Afrāsiāb ‘to Afrāsiāb’, appear in the Florence manuscript dated 1217-18 and have entered Djalal Khaleghi-Motlagh’s edition. Except for bad instead of ba/e, these are dialectal. The same is true of the works of the poets of the Ghaznavid (q.v.) period, such as Farroḵi Sistāni, ʿOnṣori (qq.v.), and others, which have almost no dialectal elements. Maqdesi states that the language of Ṭus and Nesā (Nisa, q.v.) is close to the language of Nishapur (p. 334). Therefore, the view of Parviz Khanlari (q.v.), who maintained that Persian appeared in its classical form in the 6th century (Ḵānlari, I, pp. 269-75), seems to be incorrect. This language was separated from the languages of various regions of Khorasan, which had their own dialectal forms or possibly early forms no longer current in classical Persian, and was not the language of any particular region—just as official Middle Persian was also not the language of any particular region, and had, several centuries earlier, become separated from the Middle Persian current in Fārs, which later acquired the status of a dialect.
One can reasonably suppose that the Khorasani writers who wrote in classical Persian spoke a local form of Persian in their own towns or villages. Here, we will not be concerned with describing classical Persian, which has, with some changes, survived to the present time, but will only study the characteristics of texts whose authors were influenced by their own local form of Persian. From among the huge number of such texts, we will only discuss those written in two major cities/regions of Khorasan: Nishapur and Herat (qq.v.). Then we will consider various types of Persian current in several Khorasan regions at the present time. However, it should be noted that Asadi Ṭusi (q.v., d. 465/1072-73) collected in his Loḡat-e fors early and dialectal words and forms found in the poetry of the Samanid and Ghaznavid periods. Some of these words were borrowed from Sogdian, such as naḡz ‘elegant, precious’; faḡ ‘idol’; tart-o mart ‘ruined, useless’; and others, which were pointed out by Henning (1939, p. 94). Some others, with the consonant l, came from Balḵ, such as linj ‘to draw out’; alfaγdan for anduḵtan ‘to accumulate’; and mol ‘wine’ (see Lurje and Yakubovich, 2017, pp. 319-41). Some others were taken from other eastern Iranian languages, such as espaγōl or aspγōl, meaning esparza and bazrqaṭunā ‘fleawort’, which the dictionaries define as “horse’s ears,” and which was taken from one of the eastern languages, such as Sangliči or Eškāšemi (q.v.), because guš ‘ear’ is pronounced γol̥ in Sangliči and γol in Eškāšemi (Morgenstierne II, p. 394).
Characteristics of Nishapuri Persian. The first person to point out the characteristics of Nishapuri Persian was Maqdesi (p. 334). He states that the language of Nishapur is “eloquent and easily understandable,” except that at the beginning of words, that is, in the imperative form, they put a kasra and add y (=ē) as in bē-gō ‘say’ and bē-šaw ‘go’ and add a useless s to present perfect verb forms, as in beḵᵛard-ast-i for ḵᵛarda-i ‘you have eaten’; be-goft-ast-i for gofta-i ‘you have said’; and be-ḵoft-ast-i for ḵofta-i ‘you have slept’.
Apparently, the only text that is specified as composed in the Nishapuri language is a do-bayti (q.v.), which, because the words are not pointed, is incomprehensible, but apparently it rhymes in the word bu, meaning bāšad ‘may be’. This do-bayti is from an anthology kept in the Marʿaši Library in Qom dated 1252 (see Afšār, p. 6). In the Asrār al-tawḥid, the word nāvona ‘a used bed-sheet’ has been recognized as clearly Nishapuri (Mayhani, p. 80). In Moḥammad b. Abi’l-Barakāt Jawhari Nišāpuri’s Jawāher-nāma (q.v.; dated 1196), several Nishapuri words have also been noted (see Neẓāmi, pp. 67, 167, etc.). From the fact that the above-mentioned do-bayti has been clearly recognized as Nishapuri, we may conclude that the language current among the inhabitants of Nishapur at the time when this poem was recited—probably in the 12th or 13th century—was different from classical Persian. On the basis of some expressions related from Abu Saʿid b. Abi’l-Ḵayr (q.v.) in the Asrār al-tawḥid (if these are indeed attributable to Nishapur and not to Abu Saʿid’s birthplace, Mayhana), there were also grammatical differences in the language of the Nishapuris. Someone asked Abu Saʿid: Ma-rā ba-paḏir ‘Accept me’. The shaikh said: Ne-t wā, meaning na-bāyad-at ‘It is not appropriate for you’; ma-t bēnamā, enšāʾ Allāh mabinamat ‘God willing, I won’t see you’ (see Mayhani, introd., p. 109, text, p. 116).
In order to acquire some knowledge of the characteristics of Nishapuri Persian, we shall look at some texts that were definitely written in Nishapur. Then we shall add to these the characteristics of other texts written in some towns and villages around Nishapur and share some linguistic features with the Nishapuri texts. These texts consist of the Tafsir by ʿAtiq Nišāpuri Surābāni (erroneously: Surābādi; d. 494/1100), published on the basis of the Torbat-e Jām manuscript as Tarjama wa qeṣṣaha-ye Qorʾān; Maydāni Nišāpuri’s (d. 518/1124) al-Sāmi fi’l-asāmi; Abu Jaʿfar Bayhaqi’s (d. 544/1150) Tāj al-maṣāder; Abu ʿAbd-Allāh Zawzani’s (d. 486/1093) Ketāb al-maṣāder; the Tafsir‑e Šonqoši; and an early partial translation of the Qorʾan published as Tarjama-i āhangin az do jozʾ-e Qorʾān. The characteristics of Surābāni’s Tafsir were briefly studied by Lazard (1963).
The suffix denoting continuous action of the verb in the past and the “unreal/irrealis mood” conditional in Surābāni is -ēḏ instead of classical Persian -ē (see Lazard, 1963, sec. 450). The same pronunciation existed in Herat (see below).
The -ē indicating the indefinite article in Surābāni is usually shortened to –e: asb-e for asb-i ‘a horse’; waqt-e for waqt-i ‘a time’; gomān-e for gomān-i ‘a belief, assumption’.
In Surābāni, the Tarjama-i āhangin, Bayhaqi, Zawzani and Maydāni, the preverb farāz is changed to fā(z) or used along with it (i.e., farāz; see Sādeqi, 2012, p. 357).
In the Tarjama-i āhangin and Maydāni the words mehin and behin are found as mēhēn and bēhēn ‘greatest’, ‘best’. The form bēhēn is also found in Zawzani (Sādeqi, 2012, p. 357).
In Surābāni, the final consonants of words following a vowel or a vowel plus consonant are usually dropped; for example, hanu for hanuz ‘still’; hazā for hazār ‘a thousand’; darnāk for dardnāk ‘painful’; sa for sar ‘head’; ba-dānas for ba-dānest ‘he/she knew’.
Initial g or v/w of Middle Persian changes to b in some words: beraviḏan for geraviḏan ‘to admire; follow; ally with’ (Surābāni; Tarjama-i āhangin; Šonqoši); binjiḏan for gonjidan ‘to be contained (in)’ (Tarjama-i āhangin); bezand for gazand ‘wound’ (Šonqoši); barzidan for varzidan ‘to work, practice’ (Bayhaqi; Zawzani).
The consonant ḏ indicating the second person plural of verbs is dropped in Surābāni: kardi for kardiḏ ‘you did’; bāši for bāšiḏ ‘you might be, you are’, etc.
In Surābāni the nasal n after the long vowel ā is sometimes dropped: išā for išān ‘they’; āšyā for āšyān ‘nest’; nā-nevisandegā for nā-nevisandegān ‘not writers’).
In Surābāni verb endings in many cases take the following forms:
| Singular 1st -ym =-ēm 2nd -y = i 3rd -d = aḏ; yḏ = ēḏ | Plural -m = -em -nd = -end or -and; -d = -eḏ -yḏ = -ēḏ |
In cases where in classical Persian the direct object takes -rā, in Surābāni in many instances this particle is not used.
After the verb goftan ‘to say’, where in classical and modern Persian the connective/relative ke is used, in this text, as in Middle Persian, kū is used.
Heravi Persian. ʿAbd-al-Raḥmān Jāmi (q.v.) states in the introductory pages to his Nafaḥāt al-ons that ʿAbd-Allāh Anṣāri (q.v.) wrote his Ṭabaqāt al-ṣufiya in the Heravi (Herati) language. In the authentic writings of Anṣāri, there are linguistic characteristics that do not appear in other contemporary texts. We have three Persian books of his: the Ṭabaqāt al-ṣufiya; part of a Qorʾān commentary (tafsir) attributed to him; and another text of his sayings and (private) prayers ( monājāt, q.v.), collected under the title Kalamāt-e Šayḵ-al-Eslām (ed. Šafiʿi Kadkani).
The Ṭabaqāt al-ṣufiya was first studied in 1923 by Vladimir Ivanow (q.v.), who noted most of its linguistic characteristics. Ivanow’s edition was based on the only manuscript known at the time, which is now held by the Asiatic Society of Bengal and is dated to 1606. Gilbert Lazard’s description (1963) is also based on Ivanow’s edition. The data for the present discussion are taken from the edition by Moḥammad Sarvar Mawlāʾi (1983), which he based on five manuscripts. In his introduction, Sarvar Mawlāʾi discussed the book’s linguistic characteristics in detail (see Anṣāri, 1983, introd., pp. 119-76). We have also taken note of Ivanow’s comments.
Characteristics of the Ṭabaqāt. The third person singular of the present tense of the verb budan ‘to be’ is –id instead of ast. The conditional present forms of this verb are also as follows: bm (=bovam), by (= bovi, bāši), nby (=nabāši), byd (= bovid, bāšid), bnd (=bovand, bāšand). In one instance, instead of id, hᵛn is used (in two manuscripts), or hᵛnᵛ (in one other manuscript; p. 157). Hᵛn is also used several times in Maybodi’s Kašf al-asrār, which is based on Anṣāri’s Tafsir (introd., p. 145), and also in Baḵš-i az Tafsir (Anṣāri, 1996, pp. 213, 229).
The forms for the simple present tense of the verb šodan ‘to become’ are: šm, for šavam; šy, for šavi; šyd, for šavid; and šnd for šavand.
The personal pronoun mā ‘we’ takes the form amā.
-ēḏ is used for -ē, the indicator for an unachieved conditional in classical Persian.
The ergative construction is used in several instances: ō-t ba če bešnāḵt, for u-rā ba-če bešnāḵti ‘by what [features, signs] did you recognize him’ (Anṣāri, 1983, p. 636); giram ke-m to be ʿelm yāft, for classical giram ke man to-rā ba-ʿelm yāftam ‘let [us] presume that I found [i.e., recognized] you by [my] knowledge’ (Anṣāri, 1983, p. 167); az donyā biāmadi wa-at marā našnāḵt, for az donyā biyāmadi va ma-rā na-šenāḵti, ‘you came from the world and did not recognize me’ (Anṣāri, 1983, p. 644). This construction had gone unnoticed by any earlier scholars of Anṣāri’s works.
Use of the verb kāmestan/kāmidan ‘to desire, to want’ before another verb denotes “to be about to” do something indicated by the verb.
The verbs davidan ‘to run’ and tāftan ‘to shine (sun), to be warm’ are used in the forms davestan and (var, dar) tā(w)/bestan (compare the use of ḵvāhistan for ḵvāstan in Baḵš-i az Tafsir; Anṣāri, 1996, pp. 131, 137).
The -ē of the indefinite article is shortened to -e: nur-e (=nur-i) tāwid ‘a light shone’; qawm-e (=qawm-i) ‘a people, group’ (Anṣāri, 1983, pp. 213, 541).
Between two vowels, or after a vowel, b is changed to v/w.
It uses ī, ē instead of in, u from Middle Persian ēd ‘this’.
It uses -ī instead of the connective ke (Anṣāri, 1983, p. 515).
It uses ke in the sense of waqt-i ke, ‘when…’ (Anṣāri, 1983, p. 631).
It uses a in the sense of āngāh ‘then’, along with a suffix pronoun, exactly as the form is used in Middle Persian, but pronounced ā: ar-aš dust yāft, aš nur yāft, var dar ṭalab be-mirad, aš šafiʿ yāft ‘If he found him as a friend, then he found a light’ (Anṣāri, 1983, p. 136).
It uses anō ‘there’ from Middle Persian ānōh ‘there’.
The vocabulary of this and other Heravi (Herati) texts has been studied in detail by ʿAli Rawāqi (2016).
Characteristics of “Baḵš-i az Tafsir”. The third person singular in the present potential tense is constructed with -ā, in the form used in Parthian: tā dar yād dārā zirakān va-ḵodāvandān-e ḵerad ‘So that the clever and the wise remember’ (Anṣāri, 1996, p. 237). In Parthian, this suffix is -ā(h). The third person singular in this tense occurs in this form in some other texts as well, among them the poetry of Mawlawi (Rumi; see Abu’l-Qāsemi, p. 63). This construction is also seen in the Asrār al-tawḥid: ʿayš ḵoš bā ‘may the feast be enjoyed’ (Mayhani, p. 224). But in this book, this suffix has also spread to the first and second persons singular: Mat binamā ba morād rasida-i ‘I hope to see you not having obtained your wish’ (Mayhani, p. 295); Hič kār-rā mašāʾiā ‘I hope you were suited for no work’ (Mayhani, p. 302); Nābinā gardiā ‘May you turn blind’ (Mayhani, p. 247). This construction is also seen in the Persian Psalms, as Sims-Williams has noted (p. 368). In Moḥammad Rāzi’s al-Moʿjam, the alef (ā) at the end of these forms is considered the alef-e doʿā ‘the alef of prayer, wish’ (Rāzi, p. 155).
The indicator of the third person singular present tense takes the form -ēd: ḵᵛāhid for ḵᵛāhad ‘he wants’ and namāʾid for namāyad ‘it shows’ (Anṣāri, 1996, pp. 234, 254).
The negative particle na- is used before the auxiliary verb, as in dānesta naʾid for nadānastaʾid ‘you didn’t know’ (ibid., p. 209).
It uses tā ‘so that, until’ in the sense of yā ‘or’, from Old Persian yatā.
It uses yām for yā, from Middle Persian ayāb ‘or’, changing the final b to m. Yām is also used in several other texts, as in Qomi’s Tarjama-ye Tāriḵ-e Qom (p. 11), Ḵᵛābgozāri (p. 52) and, repeatedly, in Ṭusi’s ʿAjāʾeb al-maḵluqāt.
Došvār ‘difficult’ is pronounced dežvār, which is the Parthian pronunciation.
Viža is used in the form oviža ‘special’ (Anṣāri, 1996, pp. 140, 185, 290).
Man ‘I’ is vowelled men.
Characteristics of “Kalamāt-e Šayḵ-al-Eslām.” This text shows virtually the same features as the other two texts attributed to Anṣāri, such as the ergative construction (Anṣāri, 2015, p. 249); –i is used instead of relative ke (pp. 256, 258, etc.); –i instead of the kasra of the eżāfa (ibid., pp. 232, 233); ke instead of har ke ‘whoever’ (pp. 198, 213, 283); shortening of -ē of the indefinite particle to -e (pp. 211, 217, etc.); kāmestan used in the sense of nazdik budan ‘being near, about to’ (p. 252); the use of early forms, such as joḏ for joz ‘except’ (pp. 226, 228, etc.); amā for mā ‘we’ (p. 256); i (=ē) meaning in ‘this’ (p. 243); dežvāri for došvāri ‘difficulty’ (p. 253); ezdudan for zedudan ‘to rub off’ (pp. 235, 236, 267). Dialectal forms also occur: šenāḡ, šenā ‘swimming’, in place of šināw (=senāḡ) (p. 219); bēhāna for behāna ‘figurative’ or ‘invalid’ (as opposed to ḥaqq ‘truth’); زﭬﺎن = zavān or zafān ‘language, tongue’ (p. 182), but in other instances in the forms zavān or zafān; bandaʾi for bandagi ‘servitude’; amāl for hamāl ‘equal’ (pp. 220, 250, 255); ayna (ʾynh) for āyena ‘mirror’ (pp. 259, 264).
Ḵaryā (ḵryʾ) is used for ḵaridār ‘buyer’ (p. 281; also in the Ṭabaqāt, Anṣāri, 1983, p. 412); āguš for āḡuš ‘bosom’ (p. 284); aniz, repeatedly for niz ‘also’; f-, in place of b-, (pp. 210, 211, 257); far for bar ‘upon’ (p. 215); fā, pā, along with farā; vā (=bā) ‘with’ (p. 231); -em, for -ēm, the indicator of the first person plural: agar az dustānem ḵašyat az miān bardār va agar az mehmānānem nikumān dār ‘If we are friends, put fear aside, and if we are guests, treat us with respect’. In many words, b between two vowels is changed to v or w.
The characteristics of modern Heravi/Herati Persian are discussed by Moḥammad Āṣef Fekrat. On the basis of his research, two historical vowels, ō and ē, are still used in Herat. The endings of verbs in the simple past are: -om, -i, -ø, -ēm, -ēm, and -am. The future tense is formed by the addition of the unconjugated verb ḵā to the beginning of the simple past forms (e.g., … ḵā-raftom, etc.). At the end of the third person singular of the simple past, just as in some other Iranian Khorasani languages, the suffix -ak is added (e.g., goftak, for goft ‘he said’). The progressive aspect is formed by adding the particle hay or hay hay in an unchangeable form before the simple present and the past progressive tenses: hay menwesēm (=dārim minevisim) ‘we are writing’; hay hay mēzad ‘he kept on striking’.
Today, dialects and local forms of Persian are used in different regions of Khorasan; some of these have distinctive grammatical constructions not taken from classical Persian. Here will be discussed the language of three towns for which relatively good descriptions are available.
Persian current in Nishapur. We have no description of Nishapuri Persian; but we have more information about Sabzavār, which neighbors Nishapur. We have good descriptions of the village of Boruḡan, from the rural district of Kāh in Bāštin section from the Dāvarzan subprovince, which is not that far from Sabzavār. Apparently the language of this village differs little from that of the regions around Sabzavār. In Boruḡan, the two vowels /ō/ and /ē/ still exist, and create some contrast (see Boruḡani). The compound consonant /ḵᵛ/ also still exists in some words. The group of diphthongs creates contrast: /aw/ with /ow/ and /ay/ with /ey/. In some words, the guttural consonant ḥ occurs in words of Arabic origin such as haykal and also in Persian words such as handaq ‘moat, trench’ (Ar. ḵandaq): thus, ḥaykal, ḥandaq.
At the beginning of verb forms in the simple past, the present perfect, and the progressive pluperfect, the preverbs be/-bo are added: beraftom for raftam ‘I went’; boḵordēyom for ḵorda-am ‘I have eaten’; beḵereye-biyom for ḵarida budam ‘I had bought’. With the addition of the negative indicator na- to these forms, the preverb/prefix be/bo is dropped; but in the village of Češām (former Češom), the negative indicator follows be/bo: benaref for naraft ‘he didn’t go’. In the imperative or the conditional in the sense of the imperative, na- follows o /be: benaškeni for naškani ‘don’t break (it)’. In the third person singular of the simple past, as in Herati/Heravi Persian as well as some other varieties of Khorasani Persian, including Qučān, the suffix -ek (-ak in Qučān) is added: goftek for goft ‘said’.
In the future tense also, as in Herati, the word ḵa, from the verb ḵᵛāstan, is placed before the conjugated forms of the verb in question: ḵa-reftom for ḵvāham raft, ḵa refti for ḵvāhi raft, etc. This construction also occurs in Sabzavar and other villages, but is becoming less frequent. In the case of the negative, the negative indicator na- comes before ḵa: na-ḵa reftom for na-ḵᵛāham raft ‘I will not go’.
Preverbs are widely used: vā as in vāndōḵtan, for andāḵtan va pahn kardan ‘to spread out, for example a tablecloth or bedding’; ver, as in verkešeyan for kašidan ‘to pull’; de/di as in debistan for bastan ‘tie, bind’; digziyan for gozidan ‘to choose’; foru, as in foru-kuftan ‘to fight with someone and injure him’; de, as in de-kuftan ‘to press one’s foot on, turn around, twist’.
The first part of the word howsā ‘there’ is probably borrowed from Parthian hō.
A detailed lexicon of the dialect of the town of Sabzavār has been compiled by Ḥasan Moḥtašam (1996). The book’s introduction contains a brief, sixteen-page discussion of the differences between Sabzavāri and “official” Persian. Some of these differences, which also differ from what has been written above, are as follows: In Sabzavāri Persian, there is an initial consonant cluster produced by dropping the vowel between two consonants, e.g., psar for pesar ‘son, boy’; bča for bačča ‘child’; plašt for palašt ‘impure’.
The vowel ē changes to e, as in meḵ, for mēḵ ‘nail’; sev for sēv ‘apple’; bed for bēd ‘willow’; zer for zēr ‘beneath’. But ō has changed to u (modern): rud for classical rōd ‘river’; pust for classical pōst ‘skin, hide’.
The vowel ū has changed to ī (modern i): ʿaris, tābit, ami, for ʿarus ‘bride’; tābut ‘a bier, coffin’; and amu ‘a paternal uncle; churn’.
Ḵᵛa is changed to ḵā, as in noḵād for noḵod ‘bean’; ḵārden for ḵordan ‘to eat’. Apparently the intermediary stage in this development was ḵō, because the words rowšan ‘bright’ and rowqan ‘oil’ are pronounced rāšan and rāqan. Rōbāh ‘fox’ becomes rāvā; kahwan becomes kōhan, then kāhan ‘old’.
Ān and ām have changed to on and om: meydon for meydān ‘square’; non for nān ‘bread’; bādom for bādām ‘almond’. In some words, the n has been dropped and the ā changed to –u: āvizu for āvizān ‘hanging’; resmu for rismān ‘rope’; miyu for miān ‘middle’.
The vowel ā in the initial syllable of a word usually becomes e, and sometimes a: evešu for āvišan ‘oregano’; zeni for zānu ‘knee’; beleš for bāleš ‘pillow’; jerow for jāru ‘broom’. The change of ā in the initial syllable to a usually appears in forms in which the vowel in the second syllable is already a: ama for āmad ‘he came’; ataš for ātaš ‘fire’; ḵana for ḵāna ‘house’; ahan for āhan ‘iron’.
Instead of the (attributive) suffix -i, -u is used: ḵemiru for ḵamiri ‘doughy’; česbu for časbi ‘sticky’; noḵādu for noḵodi, ‘related to peas’.
The simple past, present perfect, past perfect and past conditional are used with the preverb be-: beḵārdom for ḵordam, beḵārdem for ḵordim, beḵārde for ḵordid. The future is formed by adding ḵa, from the verb ḵvāstan, at the beginning of the conjugated forms of the verb: ḵa-ḵārdom for ḵᵛāham ḵord, ḵa-ḵārdem for ḵᵛāhim
ḵord, ḵa-ḵārde for ḵᵛāhid ḵord.
Persian current in Qāʾen. There is a detailed linguistic description for Qāʾen, a city in southern Khorasan, by Reżā Zomorrodiān (1989). Qāʾeni consonants are like Persian, except that ž does not exist. The historical vowels ō and ē still exist in Qāʾeni. There is one other vowel which does not exist in standard Persian, that is, open ԑ, as opposed to /e/, /ԑ̄/, and /a/. There are also the diphthongs /au/ and /ai/, in contrast to /ou/ and /ei/.
In Qāʾeni, the preverbs b-/be/bo /bi are also added at the beginning of the simple past, present perfect, past perfect and imperative forms: bo-ḵārdom for ḵordam ‘I ate’; be-baft ԑ yom for bāfta-am ‘I have woven’; b-um ԑ d ԑ yem for āmada budim ‘we had come’; bi-y-āyei for be-yāʾid ‘Come! [pl.]’.
The future tense is constructed in the three persons singular by adding be/bo to the conjugated forms of the verb ḵᵛāstan and attaching the non-changeable form of the main verb: bo-ḵom-bāf for ḵᵛāham bāft, be-ḵei-bāf for ḵᵛāhi bāft, be-ḵ ԑ bāf for ḵᵛāhad bāft ‘he/she will weave, etc.’. But in all three persons plural the verb ḵᵛāstan does not change, while the main verb in question is conjugated: be-ḵ ԑ -bāftem for ḵᵛāhim bāft, be-ḵ ԑ -bāftei for ḵᵛāhid bāft, be-ḵ ԑ -bāftan for ḵᵛāhand bāft.
Preverbs are also in use, as in va/vā: va-dada for dādan; va-j ԑ sta for jastan, jahidan ‘to jump’. In the case of the conditional and imperative modes and future tense this preverb takes the form vā; in the remaining tenses and modes it is pronounced va.
The preverbs v ԑ r and d ԑ are also used: V ԑ r: v ԑ r-ḵ ԑ sta for barḵāstan ‘to rise’; v ԑ r-gofta for goftan ‘to say’; d ԑ: d ԑ -b ԑ sta for bastan ‘to close’; d ԑ -ḵ ԑ zida for dar ḵazidan ‘to hide; creep into a corner’.
The preposition h ԑ corresponds to hā in Māzandarāni and central dialects; it is itself derived from frā, the shortened form of Middle Persian frāz ‘forth, forward’, corresponding to classical Persian farā(z). This preposition has two meanings: (1) it indicates direction, as in h ԑ y bāy dada, for ba-bād dādan, farā bād dādan ‘to cast to the winds; to waste’; (2) it denotes being in a certain place, as in h ԑ tu ḵun ԑ, for dar ḵāna ‘in the house, at home’.
There is also one postposition in Qāʾeni that does not exist in Persian: -eg ԑ, which conveys the large size of the basic word: sibeg ԑ ‘large apple’; sageg ԑ ‘large dog’. Zomorrodiān has also published a lexicon of Qāʾeni (2006).
Injā ‘here’ and ānjā ‘there’ are expressed heiǰā and houǰā. Evidently the first word is constructed from ē+jā and the second from ho+jā; ē has been taken from Middle Persian ēd, “this”, hou from Parthian hō ‘that’.
Persian current in Birjand. Birjand (q.v.) is south of Qāʾen and is in fact the southernmost town of Khorasan. There is an old, brief description of Birjandi Persian by Vladimir Ivanow (1928, pp. 246-55), but this description is no longer relevant. Jamāl Reżāʾi published a complete and precise description of Birjandi Persian in 1998. According to Reżāʾi, Birjandi consonants are the same as Persian, but without ž. There are three long vowels, ō, ā, and ē, in Birjandi, and one other vowel, /ə/. This last vowel is used as the final vowel of a phrase or a single word, and is an allophone of /a/ or /e/; for example, ḵordə for ḵordan ‘to eat’; pardə for parda ‘curtain’. The vowel /ē/ is a continuation of the older usage of /ē/; but this is not the case for /ō/, which is taken either from early /ā/ preceding /n/ and /m/, as in jōm for jām ‘cup, goblet’ and nō(n) for nān ‘bread’, or following oh, as in zōr for ẓohr ‘noon’.
Verbs, whether intransitive or transitive, are conjugated in two ways. The first construction follows the ergative construction of Middle Persian, either by using the attached pronominal suffix after the vowels o or e, or by bringing the separate pronouns before the verb, which in both cases is constructed with the unconjugated form of the verb, as in: om gof(t) (=goftam ‘I said’), et gof(t) (=gofti), eš gof(t) (=goft), mā gof(t) (=goftim), tu gof(t) (=goftid), šu gof(t) (=goftand); and om honšast (=nešastam ‘I sat’), etc.
In the case of such forms as mo gof(t), to gof(t), u gof(t), mā gof(t), šemā gof(t), unō gof(t), mo honšas(t), in all these conjugations the prefix be/bo usually precedes the verb stem, as in: om beraf(t), mo beraf(t), unless the verb is preceded by a preverb, in which case be/bo is not used.
The second construction is exactly the same as in Persian, except that the prefix be/bo is added to all of the conjugated forms, as in berafton, bebordom.
In the present perfect, the prefix be/bo is added to all forms in all conjugations.
The past continuous is formed by adding the continuous preverb ma/me to all forms.
The future tense has two conjugations: (1) The verb ḵᵛāstan, with the prefix o-/be, is conjugated, while the main verb is unchanged, as in beḵom raf(t) for ḵᵛāham raft; (2) the auxiliary verb remains unchanged, while the main verb is conjugated, as in beḵā raftom for ḵᵛāham raft, beḵā rafti for ḵᵛāhi raft, and so on.
The negative sign ne/a comes after the prefix be: be-ne-yārdom for nayāvardam ‘I did not bring’; da-na-basti for nabasti ‘you [sg.] did not close’.
In the case of numerals, an interesting point is that in counting from twenty upwards the smaller number comes before the larger: yakbist for twenty-one, nōsi for thirty-nine, and so on.
In the case of pronouns, worth mentioning is the use of -de in place of the attached pronoun for the third person singular when it is an object: ma-barem-de for mi-barim-aš ‘we carry him/her/it’. The adverb inje for injā ‘here’ is also used in place of the third person singular pronominal suffix when it is a possessive pronoun: dar-inje for dar-aš ‘its door’; šāḵ-inje for šāḵ-aš ‘its branch; its horn’.
As for suffixes, there is only one not used in Persian: -eyə, indicating honor, esteem, nobility, wonder, and occasionally belittling: xar-eyə, for ḵar-e bozorg ‘large donkey’; deraḵt-eyə for deraḵt-e bozorg ‘large tree’. This is comparable to the suffix –egε in Qāʾeni.
In attaching the -i of the gerund to words ending in ə, the intervening consonant g is not used, and the vowels ə and -i are changed to -ey: tošney for tešnagi ‘thirst’; zendey for zendagi ‘life’; and so on.
Jamāl Reżāʾi has also published a detailed lexicon of Birjandi (1994).
There are also descriptions of the Persian current in other cities in Khorasan, including Esfarāyen (q.v.) and Ṭabas; and a good number of lexicons have been published for various cities.
Bibliography
- Moḥsen Abu’l-Qāsemi, “Feʿl-e doʿāʾi dar zabān-e fārsi,” in ʿAli-Ašraf Ṣādeqi, ed., Yād-nāma-ye Doktor Aḥmad Tafażżoli, Tehran, 2000, pp. 61-66.
- Iraj Afšār, “Čand soruda-ye ṭabari, nisāburi, va bahlavi,” Guyeš-šenāsi 2, 2004, pp. 4-6.
- Friedrich C. Andreas and W. B. Henning, “Mitteliranische Manichaica aus Chinesisch-Turkestan: II,” Sitzungsberichte der Preussischen Akademie der Wissenschaften. Philosophisch-historische Klasse, 1933, pp. 294-363.
- ʿAli b. Aḥmad Asadi Ṭusi, Loḡat-e fors, ed. F. Mojtabāʾi and ʿA-A. Ṣādeqi, Tehran, 1986.
- ʿAbd-Allāh Anṣāri, Ṭabaqāt al-ṣufiya, ed. Moḥammad Sarvar Mawlāʾi, Tehran, 1983.
- Idem, Baḵš-i az tafsir-i kohan be-Pārsi, ed. Mortażā Āyat-Allāhzāda Širāzi, Tehran, 1996.
- Idem, “Kalamāt-e Šayḵ al-Eslām,” ed. Moḥammad-Reżā Šafiʿi Kadkani, Hargez va hamiša-ye ensān, Tehran, 2015.
- Abu’l-Fażl Boruḡani, Barrasi-e zabān-šenāsāna-ye guyeš-e Sabzavār, Sabzavar, 2002.
- Arthur Christensen, Contributions à la dialectologie iranienne, 2 vols., Copenhagen, 1930-35.
- Ebn al-Nadim, Ketāb al-fehrest, ed. Reżā Tajaddod, Tehran, 1971; tr. Bayard Dodge, as The Fihrist of al-Nadim, 2 vols., New York, 1970.
- Moḥammad Āṣef Fekrat, Fārsi-e Heravi, Mashhad, 1997.
- Moḥammad Ḡazāli, Kimiā-ye saʿādat, ed. Ḥosayn Ḵadivjam, Tehran, 1982.
- Idem, Naṣiḥat al-moluk, ed. Jalāl Homāʾi, Tehran, 1972.
- Antoine Ghilain, Essai sur la langue parthe: son système verbal d’après les textes manichéens du Turkestan Oriental, Leuven, 1939; repr. 1966.
- W. B. Henning, “Mitteliranisch,” in B. Spuler, ed., Handbuch der Orientalistik, Erste Abteilung, IV/1, Leiden, 1958, pp. 20-129.
- Idem, “Sogdian Loan-words in New Persian,” BSOS 10, 1939, pp. 93-106; repr. in Selected Papers I, Acta Iranica 14, 1977, pp. 639-52.
- Yuliĭ A. Ioannesyan, Geratskiĭ dialekt yazyka dari sovremennogo Afganistana (Herat dialect of the Dari language of modern Afghanistan), Moscow, 1999.
- Idem, “Situating the Khorasani Dialects within the Persian-Dari-Tajiki Linguistic Continuum,” in Franklin Lewis and Sunil Sharma, eds., The Necklace of the Pleiades: Studies in Persian Literature Presented to H. Moayyad on His 80th Birthday, Leiden, 2010, pp. 267-78.
- Vladimir Ivanow, “Tabaqát of Ansári in the Old Language of Herat,” JRAS 1, 1923, pp. 1-34, 337-82.
- Idem, “Persian as Spoken in Birjand,” JRASB, N.S., 14, 1928, pp. 235-351.
- ʿAbd-al-Raḥman Jāmi, Nafaḥat al-ons men ḥażarāt al-qods, ed. Maḥmud ʿĀbedi, Tehran, 1991.
- Ḵᵛābgozāri, ed. Iraj Afšār, Tehran, 1967.
- Parviz Nātel Ḵānlari, Tāriḵ-e zabān-e farsi, 4 vols. in 5, repr. Tehran, 1986.
- Abu ʿAbd-Allāh Moḥammad Ḵᵛārazmi, Mafātiḥ al-ʿolum, Leiden, 1895.
- Gilbert Lazard, La langue des plus anciens monuments de la prose persane, Paris, 1963.
- Idem, “Pahlavi, Pârsi, Dari: les langues de l’Iran d’après Ibn al-Muqaffaʿ,” in C. E. Bosworth, ed., Iran and Islam in Memory of the Late Vladimir Minorsky, Edinburgh, 1971, pp. 361-91.
- Idem, “Remarques sur le Fragment Judéo-Persane de Dandān Uiliq,” A Green Leaf: Papers in Honour of Jes P. Asmussen, Acta Iranica 28, Leiden, 1988, pp. 205-9.
- Idem, “La dialectologie du persan préclassique à la lumière des nouvelles données,” Studia Iranica 43, 2014, pp. 83-97.
- Wolfgang Lenz, “Die nordiranischen Elemente in der Neupersischen Literatursprache bei Firdosi,” Zeitschrift für Iranistik und Indologie 4, 1926, pp. 251-316.
- vel Lurje and Ilya Yakubovich, “The Myth of Sogdian Lambdacism,” in Zur lichten Heimat: Studien zu Manichäismus, Iranistik und Zentralasienkunde im Gedenken an Werner Sundermann, Iranica 25, Wiesbaden, 2017, pp. 319-41.
- David Neil MacKenzie, “An Early Jewish-Persian Argument,” BSOAS 31/2, 1968, pp. 249-69.
- Abu ʿAbd-Allāh Moḥammad Maqdesi (or Moqaddesi), Ketāb aḥsan al-taqāsim fi maʿrifat al-aqālim, ed. M. J. de Goeje, Leiden, 1906.
- Josef Marquart (Markwart), Erānšahr nach der Geographie des Ps. Moses Xorenacʿi, Abhandlungen der Königlichen Gesellschaft der Wissenschaften zu Göttingen, phil.-hist. KL. N.F. 3/2, Berlin, 1901.
- Moḥammad b. Monawwar Mayhani, Asrār al-tawḥid fi maqāmāt al-Šayḵ Abi Saʿid, ed. Moḥammad-Reżā Šafiʿi Kadkani, 2 vols., Tehran, 1987.
- Ḥasan Moḥtašam, Farhang-nāma-ye bumi-ye Sabzavār, Sabzavar, 1996.
- Davoud Monchi-Zadeh, Wörter aus Xurāsān und ihre Herkunft, Leiden, 1990.
- Georg Morgenstierne, Indo-Iranian Frontier Languages II, Oslo, 1938.
- Friedrich Müller, “Ein syrisch-neupersisches Psalmenbruchstück aus Chinesisch-Turkistan,” in Gotthold Weil, ed., Festschrift Eduard Sachau, zum siebzigsten Geburtstage gewidmet von Freunden und Schülern, Berlin, 1915, pp. 215-22.
- Neẓām-al-Molk, Siar al-moluk, ed. Herbert Darke, Tehran, 1968.
- Moḥammad b. Abi’l-Barakāt Jawhari Nišāburi, Jawāher-nāma-ye Neẓāmi, ed. Iraj Afšār and Moḥammad Rasul Daryāgašt, Tehran, 2004.
- Ludwig Paul, A Grammar of Early Judaeo-Persian, Wiesbaden, 2013.
- Ḥasan b. Moḥammad Qomi, Tāriḵ-e Qom, ed. Jalāl-al-Din Ṭehrāni, Tehran, 1934. ʿAli Rawāqi, Guna-šenāsi-ye matnhā-ye fārsi-e: Guna-ye fārsi-ye heravi, Tehran, 2016.
- Šams-al-Din Moḥammad b. Qays Rāzi, al-Moʿjam fi maʿāyir ašʿār al-ʿajam, ed. Moḥammad Qazvini and M.-T. Modarres Rażawi, Tehran, 1935.
- Jamāl Reżāʾi, Vāža-nāma-ye guyeš-e Birjand, Tehran, 1994.
- Idem, Barrasi-e guyeš-e Birjand, Tehran, 1998.
- ʿAli Ašraf Ṣādeqi, Takwin-e zabān-e fārsi, Tehran, 1978.
- Idem, “Baḥṯ-i dar bāb-e ketāb-e Pol-i miān-e šeʿr-e hejāʾi wa-ʿarużi…,” in Ḵerad bar sar-e jān, ed. Moḥammad Jaʿfar Yāḥaqqi et al., Tehran, 2012, pp. 349-88.
- Idem, “Kasra-ye eżāfa ba-jā-ye ke-ye mawṣul,” Farhang-nevisi 11, 2016, pp. 3-7.
- Nicholas Sims-Williams, “Early New Persian in Syriac Script,” BSOAS 74, 2011, pp. 333-74.
- Werner Sundermann, “Einige Bemerkungen zum syrisch-Neupersischen Psalmenbruchstuck aus Chinesisch-Turkestan,” in Ph. Gignoux and A. Tafazzoli, eds., Mémorial Jean de Menasce, Louvain, 1974, pp. 441-452.
- Idem, “Mittelpersich,” in Rüdiger Schmitt, ed., Compendium Linguarum Iranicarum, Wiesbaden, 1989, pp. 138-64.
- Abu Bakr ʿAtiq Nišāburi Surābāni, Tarjama wa qeṣṣahā-ye Qorʾān, ed. Yaḥyā Mahdawi and Mahdi Bayāni, Tehran, 1959.
- Tafsir-e šonqoši, ed. Moḥammad-Jaʿfar Yāḥaqqi, Tehran, 1976.
- Tarjama-i-e āhangin az do jozʾ-e Qorʾān-e majid, ed. Aḥmad ʿAli Rajāʾi, as Pol-i miān-e šeʿr-e hejāʾi va ʿarużi-e fārsi dar qorun-e avval-e hejri, Tehran, 1974.
- Moḥammad b. Maḥmud b. Aḥmad Ṭusi, ʿAjāʾeb al-maḵluqāt wa ḡarāʾeb al-mawjudāt, ed. M. Sotūda, Tehran, 1966.
- Bo Utas, “The Jewish-Persian Fragment from Dandān-Uiliq,” Orientalia Suecana 27, 1968, pp. 123-36.
- Ṣafi-al-Din Wāʿeẓ Balḵi, Fażāʾel-e Balḵ, tr. ʿAbd-Allāh Moḥammad-Ḥosayni Balḵi, ed. ʿAbd-al-Hayy Ḥabibi, Tehran, 1971.
- Reżā Zomorrodiān, Barrasi-ye guyeš-e Qāʾen, Mashhad, 1989.
- Idem, Vāža-nāma-ye guyeš-e Qāʾen, Tehran, 2006.
- Yutaka Yoshida, “Some New Interpretations of the Two Judeo-Persian Letters from Khotan,” in Almut Hintze, Desmond Durkin-Meisterernst, and Claudius Naumann eds., A Thousand Judgements: Festschrift for Maria Macuch, Wiesbaden, 2019, pp. 384-94.
- Zhang Zhan 張湛 and Shi Guang 時光, “Yijian xin faxian Youtai Bosiyu xinzha de duandai yu shidu 一件新發現猶太波斯語信箚的斷代與釋讀” (Dating and interpretation of a newly discovered Judaeo-Persian letter), in Dunhuang Tulufan yanjiu 敦煌吐魯番研究 (Journal of Dunhuang and Turfan Studies) 11, 2008, pp. 71-99, 1 color plate.