the terms dialect and language overlap; in general, language refers to the more or less unified system of the phonology, grammar, and lexicon that is shared by the speakers of a country, or geographic region, or a socially defined group, whereas dialect (Pers. lahja, gūyeš) focuses on varieties of a language.


DIALECTOLOGY. Introduction. The terms dialect and language overlap. In general, language refers to the more or less unified system of the phonology, grammar, and lexicon that is shared by the speakers of a country, or geographic region, or a socially defined group, whereas dialect (Pers. lahja, gūyeš) focuses on varieties of a language. In that sense, any dialect can be considered a language, and vice versa. In popular usage, dialect also refers to a speaker’s accent (lahja, but not gūyeš), i.e., peculiarities in his pronunciation, including stress and pitch (such as a German accent, or the noticeable “melody” of Isfahani Persian vs. that of Tehran).

More commonly, dialect refers to groups that are noticeably different grammatically, phonologically, and lexically. They may be either closely related varieties of the same language (such as Khorasani vs. Tehrani Persian, or the Persian spoken in Persia proper vs. the variety spoken in Afghanistan, Darī, vs. Tajiki Persian) or of more distantly related languages (such as Kurdish vs. Persian, or Pamir dialects vs. Pashto). It is not always possible to distinguish whether a dialect is a variety of one language or of two closely related languages. Mutual intelligibility is one criterion, but intellibility is very much a factor of linguistic and cultural exposure. Social identity often overrides linguistically defined dialectal relationships. Speakers of socially lower status tend to identify themselves with those of higher status.

The recognition of linguistic and dialectal differences by indigenous writers and travellers alike throughout Persian history is well known. Early writers remark on the diversity of languages, including non-Iranian languages preserved among Iranians. Lōḡat-e fors quotes mainly eastern poets at a time when New Persian had already established itself as a literary language. Regional dialects of Persian were observed, most notably by geographers (e.g., Eṣṭaḵrī, pp. 91, 167, 314; Ebn Ḥawqal, pp. 254, 348, 490; Maqdesī, pp. 334-35, 368, 378, 398, 418), often with evaluative comments as to their elegance, coarseness, etc. Dictionaries of Persian began to record the geographic origin of certain words.

As to the dialects of the New Iranian period, important sources are secular and religious texts in local and regional dialects, written since the early centuries of Islam. Those texts, as well as the dialects of local religious minorities, mainly Zoroastrian and Jewish, preserve the local tongues where Persian or Turkic have eliminated them (see the various bibliographical studies by Afšār and the overview of dialectological studies in Iran by Yarshater, 1970). On the other hand, important especially for the study of the pre-Islamic diversity is the so-called Nebenüberlieferung, i.e., Iranian words and loans in non-Iranian languages, most notably Armenian (see Schmitt, 1989).

Dialectology as an academic subfield of the discipline of linguistics developed in the 19th century, together with the comparative-historical study of Indo-European, to which especially the study of Old Iranian contributed considerably. (Note the early distinction of the IE languages into a western centum-group and an eastern satəm-group, epitomized by the reflexes of IE velar palatals, *ḱṇtom “100” in Latin and Avestan, respectively.) The second half of that century was also the beginning of the systematic study of pre-Islamic Iranian texts as well as systematic field work and study of the modern spoken Iranian languages and dialects.

Dialectology is essentially comparative. It has the objective of identifying linguistic relationships in geographic, historical, and social space. The comparative objective involves the study of the two main forces of dialectical divergence and convergence; that is, on the one hand, the retention, loss, and innovation of linguistic features, and their diffusion both internally throughout the lexicon, phonology, and grammar and externally, i.e., diffusion by social and geographic contact. On the other hand, it involves the study of groupings, mostly in terms of geography and history, by the identification of bundlings of isoglosses, i.e., overlapping patterns of lines of shared differentiation, either innovative or conservative. These coinciding objectives reflect the two main original approaches of comparative-historical linguistics, namely the genealogic Stammbaum theory and the “wave” theory, the former assuming a single origin and subsequent splits, the latter multiple sources.

The latter also leads to the identification of a Sprachbund, i.e., the development of similarities shared by areally adjacent or symbiotic dialects and languages, whether genetically related or not. Diachronically, this may involve reflexes of an extinct local languag, of an incoming group, or of a superimposed dominant language. The comparative-historical approach to dialectology has more recently been extented to that of linguistic typology, i.e., the systematic study of isoglosses and more importantly of phonological, morphological, and syntactic (sub-)systems or types, irrespective of genetic relationship.

These approaches are historical in the sense that they imply the reconstruction of the diachrony of change and the temporal sequence of changes. (Note the observations by biochemists that their reconstruction of the tree of diversification of the DNA of all humankind closely resembles the one reconstructed by linguists for all human languages; see Cavalli-Sforza.) The dating of such changes, specifically their major diachronic groupings into Old, Middle, and New and/or Modern, is, however, largely a crutch, based on major political events. Old Persian refers to a period between the 6th and 3rd centuries B.C.E. and Old English to a period from the Germanic invasion of the Isles to about 1,100 C.E. Many so-called Middle Iranian languages continued in one or another form, medium, and function well after the coming of Islam. The terminology here is also in part correlated with the extant text corpuses, not only chronology/history. (Note also discontinuity in available text corpuses.) No Iranian dialect or language can be studied throughout its entire history. The lesser known dialects or languages are as a rule lumped together with the dating of the emergence of a new nationally or regionally dominant language.

Lack of change and diversion in some aspects of modern or middle dialects, even if confined to one or a few items, tended to be considered evidence for their greater “originality” and tended to attract greater attention than change.

Iranian dialectology. Basic bibliographical overview. The most recent comprehensive analysis with major attention to dialectal dynamics is the Compendium Linguarum Iranicarum (Schmitt, ed., 1989). Major earlier studies include the monumental and fundamental Grundriss der iranischen Philologie (1896-1904), Hans Reichelt(1927), the Iranistik (1958), I. M. Oranskiĭ (1960, 1963, 1975, 1979), and the largely bibliographical surveys in the Current Trends (Sebeok, ed., V-VI). Further, we have two comprehensive works edited by Rastorgueva, largely organized by topic (1975) and arranged according to historical stages and dialect groups (1979-82).

i. Important specialized studies. In terms of groupings, the most decisive studies were by Wilhem Geiger (in Grundriss), and, based on observations by Mann and ultimately Friedrich Andreas (q.v.), those by Paul Tedesco, with reference to the evidence of the Middle Iranian Turfan texts, and by Wolfgang Lentz with reference to variations in the New Persian of the Šāh-nāma. There are also studies by Harold Bailey, David MacKenzie (1961), Gernot Windfuhr (1975), and Pierre Lecoq.

ii. Phonology, morphology, and syntax. On phonology we have studies by V. S. Sokolova (I. Baluchi, Kurdish, Ṭālešī, Tātī-Persian; II. Ossetic, Yāḡnōbī, Pamir languages), and by Don Stilo on the typology of waves of palatalization affecting Iranian and non-Iranian alike in the area stretching from Persia into the Caucasus and beyond. W. L. Heston studies typological-comparative syntax of noun phrases, pronouns, passives, nominalized verb forms, and coordination and subordination in early New Persian, Middle Persian, Sogdian, and Khotanese. C. P. Masica included Iranian evidence, confined to contemporary standard Persian, Kurdish, and Baluchi, as part of his study on a small number of syntactic types across Eurasia. Hans Seiler traced relative clauses and nominal subordination from Old Iranian to Modern Persian as part of a general linguistic-typological study (see also Haider and Zwanziger). M. M. Sakhokiya’s work is a typological study of possessives, transitivity, and ergativity in Old Iranian, Old Armenian, and Old Georgian; J. R. Payne (1980) conducts a typological study of the loss of ergative constructions in the Pamir languages, and Bossong’s extensive typological study concerns the case systems of some twenty-six modern Iranian languages (see CASES). Iranian in the general context of universal syntactic typology of agency and its morphological representation is discussed in numerous specialized and general linguistic-typological studies by Gilbert Lazard (see bibliography). Windfuhr (1985) included Persian, as representative of much of Iranian, in his general model of verbal categorizations and their multidimensional interrelationships and shifts (see also Windfuhr, 1987). The category of inference is discussed by Windfuhr (1982; see also Lazard, 1985). Harald Haarmann included Persian as part of his study of this category across Eurasia.

The systematic cartographic analysis of variety and groups has so far been limited. Besides Edel’man (1968), recent less selective areal studies are the lexical maps published by the Linguistic Atlas of Afghanistan (Redard, 1974). Confined to the Central dialects, but of general import for the study of Iranian dialectology, especially in terms of its inclusion of socio-linguistic issues, is the study by Krahnke, which includes numerous maps and identifies two intersecting directions of isoglottic waves.

Groupings. Iranian languages belong to the Indo-European language family, specifically to those languages which in their documented texts identify themselves as Aryan (q.v.). There has been much recent debate about the original areas of Indo-European as a whole. In one theory Indo-European is considered one of the language groups, together with Afro-Asiatic and Dravidian, that are suggested to have originated in the Near East (Gamkrelidze and Ivanov).

The staging area of the Aryan branch of Indo-European was most likely the steppes of Central Asia, from the Caspian and Aral Seas and north. Members of one group of Aryans appear to have moved, by the beginning of the 2nd millennium B.C.E., to much of the south. Most important was the move of one group, the so-called Indo-Aryans, to the southeast, first to the area of Afghanistan and then to northwest India. The group now identified as Iranians, or Irano-Aryans, appears to have remained for some time and expanded toward the west and the east. Most decisive was the move of some of them onto the Iranian plateau. There Iranians first established rulerships, and ultimately world empires, encompassing multi-ethnic, multi-linguistic, and multi-cultural hosts. The Iranization of the area proceeded relatively slowly. The earliest staging areas for the Iranian plateau itself appear to have been in the northeast, the southwest, and the northwest. They are also the major geographically definable, even “natural,” focal points of linguistic symbiosis and turmoil (see MacKenzie, 1961; Windfuhr, 1975; and below). These appear to be the same staging areas as for Turkic since the end of the first millennium C.E. Today, Turkic and other languages have erased Iranian in the north, as Iranian languages once did to their predecessors.

The dialectological division of early Iranian is essentially based on comparative reconstruction. Iranian names of persons and places are first recorded in Mesopotamian documents of the early first millennium B.C.E. They give evidence for dialectal differentiation even at that stage. Today, Iranian languages and dialects cover an area approximately as large as that of the modern Indo-Aryan languages (see Edel’man, 1968, map 1). Ever since the earliest records, they exhibit great variety and differentiation (Schmitt, pp. 4-31; for variety in Achaemenid times, see Rossi). This variety is addressed in detail in the major recent comprehensive studies (Rastor-gueva, 1975, 1979-87; Schmitt).

Iranian languages and dialects are grouped into western and eastern categories, the former again into southwestern and northwestern, in spite of the recognition of numerous intersections at any time and in any directions. It should be noted that except for the sequence Old > Middle > New Persian, and Middle Sogdian >Yāḡnōbī, no direct predecessors or successors of the other dialects are known. This suggests greater linguistic variety at the earlier stages than that found in the documented evidence.

The incipient western division is already evident in the two attested Old Iranian languages, namely Old Persian and Avestan. This distinction is continued in documented Middle Iranian, which also shows the further distinction between West Iranian, i.e., Middle Persian and Parthian versus East Iranian, namely Sogdian, Saka (Khotanese and Tumshuqese), Choarasmian (see CHORAZMIA iii), and Bactrian. Both distinctions are continued in the modern dialects, which may be listed as follows (excluding now extinct dialects): Western Iranian: Superimposed on all is Persian, and its regional varieties, most prominently the Persian varieties in Persia, Afghanistan, and Tajikistan, and lesser known ones such as Tātī-Persian in the southeast Caucasus. The two most widely spread western dialect groups are: in the west, Kurdish dialects found in western Persia, northern Iraq and Syria, eastern Turkey, and southern Armenia; and in the east, Baluchi found in West Pakistan, eastern Persia, southern Afghanistan, and Turkmenistan. Less widely spread dialects are: In the west, Zāzā (see DIMILI) in eastern Turkey and Gōrānī near Mosul and in the central Zagros; in the northwest and the center, the broad band of Tātī and Tālešī dialects stretching from Azerbaijan to the center, to which belong Semnānī east of Tehran, and the Central dialects between Tehran, Hamadān, and Isfahan; in the north, Gīlakī on the northwest shores and Māzandarānī (formerly called Ṭabarī) on the northeast shores of the Caspian; in the southwest and south, the Lorī dialects, the dialects in Fārs, Lārestān and Baškard, and those of along the Persian Gulf coast. Pashto is the most widely spread of the eastern Iranian languages. The remaining eastern languages, each confined to small areas, include: Parāčī and Ōrmuṛī, roughly north and south of Kabul respectively; Yidḡa, Munjī, Waḵī, and the so-called Pamir languages with their own multiple subdivisions, roughly north and northeast of Kabul; Yāḡnōbī in southern Tajikistan, and Ossetic in the central Caucasus.

Alternative grouping. The term East Iranian is a misnomer, since it includes Ossetic and its predecessors in the Caucasus. As to the western division, it was recognized early on that the north-south division essentially is a reflex of the old geographic-political division between Media and Persis. This observation was revived by W. B. Henning, and more recently amply documented by Ehsan Yarshater (e.g., 1969), based on his extensive documentation of the contemporary Iranian dialects stretching from northern Azerbaijan to the center around Isfahan. That is, dialectologically, Persian and Perside dialects such as Lorī have to be recognized as representatives of a small sub-group which began to split from the remainder of West Iranian as early as the 5th century B.C.E. One therefore may speak of a division into North Iranian (instead of East Iranian), and South Iranian (instead of West Iranian), of which the southwestern dialects reflect an early regional split.

Divergence. i. Phonology. Early divergences involve prominently 1. the reflex of the old Aryan palatals, 2. fricative clusters, and 3. initial semivowels.

1. The Aryan palatals *k´/*ǵ changed over an affricate stage, i.e., */*, to s/z in common Iranian, but to θ/d > h/d in Persian, e.g., *: common Ir. das (10)> Pers. dah; *g´: common Ir. zān- (know) > Pers. dān-; also, Aryan *ḱw and *ǵw changed to sp and zb, but to s and z in OPers., e.g., *aḱw- (horse)> common Ir. asp-, but Pers. as- (NPers. savār “mounted” rider’< asa-bār-).

2. Already in Old Persian, the fricative cluster Ir. *θr had become a strident, which later changed to s, also found in Baluchi, Kurdish, and Ḵūrī. It developed to (h)r in other dialects, e.g., puθr- (son) > Pers. pes-ar, but > puhr > pūr elsewhere. This change is found extended later to other fricative clusters, in various dialects to different extent, and generally not affecting all lexemes. Both *fr and *xr >(h)r in much of the lexicon of the Central dialects, e.g., *fra-vaxš (call forth, sell) > (h)r-ōš, but Pers. for-ūš from *√wač- (to say, call) and preverb *fra- (forth); *xr-: (h)rīd vs. Pers. ḵarīd (to buy).

3. Initial semivowels *w and *y tended to remain unchanged, but in Persian and other dialects they changed to voiced stops, first *w > b or (g)w, then *y >j or ž; e.g., *w: *wāt- (wind)> wād/wāy but bād in Persian, *ward- (flower, rose) > vel, etc., but > gol in Persian (with additional late change of rd >l); *y: *yav- (barley) remains yow or in many dialects, but jow in Persian. In turn, initial voiced stops developed into fricatives in a number of eastern dialects. Also, postconsonantal semivowels tend to affect the respective consonants. This includes old initial *dw (e.g., *dwar- “door” > bar, vs. Pers. dar).

ii. Morphology and lexicon. Early divergence involves prominently: 1. pronominal suffixes and their development, 3rd singular *-šai vs. *-hai (both conditioned variants of original Indo-Ir. *-sai) > -š vs. -ē/-ī, respectively, reflected in modern dialects (e.g., in Pers. bīnī-aš vs. Ḵūrī nāk-e “nose-his”); and 2. 3rd person demonstratives, ‘this/that’, Persian and others īn/ān vs. reflexes of *im-/*aw- elsewhere (OPers. also had *im-, still fossilized in Persian em- as em-rūz/-šab/-sāl (today, tonight, this year)).

Early lexical divergence is found already in Old Iranian, e.g., Old Persian gaub- vs. Avestan wāč- (say; e.g., Pers. gū(y)-/gof-t, Sangsarī vāž-/vā-t; note the recently introduced Pers. vāža “word”). It should be noted that in regard to some of the early differences in phonology, morphology, and lexicon, the imperial Old Persian of the Achaemenid inscriptions already was no longer echtaltpersisch. Of course, there never can be a “pure” dialect.

Loss, retention, and innovation. Typologically, Old Iranian as an Indo-European language is synthetic, i.e., it is morphologically marked by a very high degree of nominal and verbal inflection, syntactically by relative free word order. The breakdown of this system is largely the consequence of phonological change, including conditioned variations and contradictions, and, more importantly, of shifts in stress patterns. Thereby inflectional endings become reduced and ultimately lost, as is the case with most of the distinctions of gender, case, and number in the nominal system, and the distinctions of tense, mood, and aspect as well as of active and passive, and the multiple derivative formations in the verb system.

Beginning as early as the 4th century B.C.E., the complex nominal case system was ultimately reduced to a morphologically unmarked system, and the verb system to a binary opposition of so-called present and preterite. This resulted initially in virtually inflection-less languages, except for the distinction of three persons in the singular and plural of the verbs and pronouns. Concomitantly, there was a process of innovations by periphrasis in the verb system, and by prepositions and postpositions in the nominal system, typologically resulting in an analytic language, with remnants of inflection distinctions.

i. Verb system. The distinctions between the categories of person, as well as those of tense and mood, in the sense of traditional grammar, were as follows:

A. Tense. The tense system inherited from Aryan consists of three sub-systems, namely present, aorist, and perfect. 1. The present system was marked by a range of formants (including zero), added to various forms of the root (e.g., present formant -a in bar-a “to carry, bear,” but formant -nau in kṛ-nau “to make, do”). Actually, this system had temporal distinctions. They were marked by differences in the personal endings, generally speaking by the addition of -i in the present tense and its absence in the past tense. In addition, the past forms tended to be marked by a so-called augment, i.e., a prefix-like marker a- (for example Old Persian present indicative baranti “they carry” vs. imperfect a-baran(t) “they carried”; cf. Old Gk. e-). 2. An aorist system developed, at least partially, as the distinction between a present and past aorist, on the model of the present system. 3. The perfect system had distinctly different endings from those of the present. It also partially developed a distinction between present, with its own endings quite dissimilar to those of the other two subsystems, and past endings, on the model of the present system.

B. Aspect. The distinctions between the inherited system of present-perfect-aorist is not one of time and tense, but one of aspect. That is, the morpho-syntactical markers express the speaker’s view of whether an action or a situation is 1. ongoing or habitual; 2. (just) done; or 3. done. It is the augment and the personal endings that indicat tense, present and past (e.g., 1. pres. kṛ-nau-ti “makes”; 2. aorist čar-t [<*kert] “made”; 3. perfect [reduplicated) ča-k(a)r-a “has made”).

C. Verbal voice. There was a further distinction of what is called “verbal voice.” This involves the distinction between active and so-called “middle” verbal voice, the former indicating the active involvement of the agent/subject, the latter indicating that the action or situation affected the agent, often translated as passive. Morphologically, this is marked in personal endings (e.g., imp. 2nd sing. active ending zero vs. middle ending -swa/-hwa in dāh “give!” vs. dāhwa “give for yourself, accept”; bara-ti “carries” vs. bara-tai” is carried”). In addition, there was the so-called passive, generally expressed by a marker -y-; there was its “opposite,” the “causative” (e.g., act. kṛ-nau-ti “makes,” pass. kṛ-ya-ti “is being made,” caus. kār-aya-ti “causes to make’).

D. Mood. In addition, the speaker could convey possibility (subjunctive forms, generally marked by -ā), suggestion (optative, generally marked by -ī; e.g., from kar-: ind. kṛ-nau-ti “he makea,” subj. kṛ-nau-a-t “he shall make,” opt. kṛ-nu-yā-t “he may make,” imp. kṛ-nu-di “he must make!” or from bar-: bar-a-ti “he carries,” bar-ā-t [<bar-a-a-t] “he shall carry,” bar-a-i-t “he may carry,” bar-a-tu “he must carry!”), and a wealth of other modalities, such as future marked by -sya (e.g., wax-šya “I will say” and sau-šya-nt- “going to save, future savior”).

The most crucial loss was that of the inherited threefold distinction of aspects, i.e., of present, aorist, and perfect, each of which had a present and a past tense. In virtually all of the modern languages/dialects, and certainly in Persian, a new ternary system of aspects has re-emerged.

Already in Old Persian and Younger Avestan the perfect had merged with the aorist into a past system. Subsequently, with the loss of the distinction between primary and secondary personal endings and the loss of the so-called augment a- marking past, the formal distinction between present and past was lost. New past/perfect forms developed. They are already found in both Old Persian and Avestan: they are based on the old perfect participle in *-ta, followed by forms of “to be”; the past agent is expressed by the oblique of the noun, pronoun, or corresponding personal suffix, e.g., imā tayat manā kṛta-māhā (this [is] what by me is made, i.e., I had made it, Mid. Pers. -mkard, be-m-ka(rd) or similar, found in most NW dialects; for origin of this construction see Pirejko, 1979). Most modern Iranian dialects still have this way of expressing past forms of the verb. It used to be called the passive construction, because in Latin or English it only can be conveyed literally in the passive mood. It now tends to be called the ergative (i.e., the “doer case”), indicating that in past tenses, the transative agent (the “doer”) is specially marked, usually by an oblique case ending. (The term was first used for a similar phenomenon in Caucasian languages, which tent to indicate the agent by a special marker, sometimes also in non-past tenses; see Boeder, 1979.) It is also widely found in ancient Near Eastern Languages such as Sumerian (see Steiner, 1979).

Typologically, these forms and their derivatives in the modern dialects resulted in the re-establishment of the ternary distinction of aspect, e.g., contemporary Persian (see Windfuhr, 1985; inferential added: Example 1).

Taking Persian as reference for Iranian in general, without assuming that these categories are pan-Iranian today, raf-t implies not only the past, “he/she went,” but occasionally a sense of the present, as for an example “there she goes,” and even the future, such as šāyad mā ham raf-t-īm “we will certainly/perhaps go too.”

The loss of the inherited Aryan past imperfective, marked by the prefix a-, is not universal. Some dialects, such as Yaḡnōbī (continuing earlier Sogdian), derive their past/preterite from the present stem. In the west, Gōrānī and Ṭālešī base their past imperfective forms on the old present stem, so the ergative construction did not affect them (although it affected the perfective and aorist); Ṭālešī has prefix a- plus present stem plus ī, Gōrānī has present stem plus ēn: Gōrānī imperf. past 3rd sing. masc. ūs-ēn-e “he was sleeping” vs. Ṭālešī and Gōrānī -ēn, both derived from earlier optative endings *-ē/-ēn, which already in OIr. could express imperfective past (e.g., Av. formal opt. -ōi-t in yavata xšyōit Yimo “as long as Yima would rule”; Y. 9.5). The Ṭālešī prefix a- may be a retention of the old augment a-.

As to other categories, dialects have developed multiple ways of differentiating their verb systems after the virtual collapse of the old system. Such development appears to be, at least partially, conditioned externally, such as through interference from Turkic. Differentiation is achieved mostly by periphrasis, which in some cases resulted in new synthetic forms. Such is the case with the widespread development of imperfect markers, serving to disambiguate this aspect from the unmarked perfective aspect. The markers were derived variously from the original temporal/locational adverbs, frequently from *hama-, (a cognate of Eng.) “same,” such as *ham-ayawa-, e.g., early New Persian hamē (contemporary NPers. -), to which functionally corresponds *hadā “same time/place,” reflected in dialectal variants de-, a(d)- (Kurdish etc.). Similarly widespread is the development of periphrastic constructions expressing Aktionsarten, such as the progressive (Eng. “I am doing”), by a great variety of means, such as verbs semantically connotating continuity or incipiency, e.g., Persian of Persia: dār-am mī-rav-am; Kabuli Persian raf-t-a mē-rav-om; Tajiki Persian raf-t-a īst-od-a am, all meaning “I am going, am about to go” (dār- “to hold, keep,” rav- “to go,” īst- “to stand, be in”), or constructions based on locative markers added to the nominalized verb, e.g., Gīlakī kara amon dar-a “he is coming” (kar “work, doing,” amo-n “coming,” dar- “in (to),” -a, 3rd sing. copula). Likewise frequent are modal constructions. These include a future marked by kām- (wish) or forms derived from “to want, wish” with various etyma. Particularly noteworthy is the frequent development of the category of inference, expressing hearsay, assumption, conclusion, and in general the distancing of a speaker, i.e., conveying that he/she did not witness an action or situation (example from cotemporary Persian literature: “Be-ṭowr-e kollī ānče az har kojā dar-yāfta-am in ast ke ostād Mākān mard-e rāzdār-ī būda . . . kamtar šūḵī mīkarda . . . hīč-kas be zendegī-e dāḵelī-e ū wāred našoda būda ast “In general what I could find out from here and there was that Ostād Mākān was a secretive man. . . He rarely made any jokes . . . Nobody had gotten through to his inner life”; ʿAlawī, pp. 13-14).

Again, there are retentions. Inflectional mood, such as the subjunctive derived from an earlier marker -ā (note the relic b-ā-d “may he be,” as in zenda b-ā-d “may he live” in contemporary Persian), was retained in a good number of dialects, as was the inflectional passive/inchoative derived from -, still preserved in early Judeo-Persian and some Persian dialects (e.g., Sangesarī: ešt-ende “he stands up” vs. ešt-i-nde “he will stand up”; Zāzā: araq-y “to drink liquor”).

ii. Nominal system (see CASES). The most serious loss was that of the distinction between nominative and accusative. These were the two morphological cases that distinguish the two most crucial syntactic cases, namely subject and direct object. Most modern dialects innovated by the grammaticalization of preposition and/or postpositions, mostly with inherent directional meaning, e.g., Persian - to mark the specific object, originally “for the sake of, concerning” to include later directional and indirect object function, or Sangsarī -de from *antar->dar “in(to).” The greatest variety is found in the eastern dialects. In addition, a more rigorous word order developed, typologically resulting in the (unmarked) order of S-O-V (subject-object-verb, compare English S-V-O). Still, many modern dialects did retain reflexes of earlier inflectional distinctions, generally indicated by the opposition of unmarked direct case and marked oblique case: sing. -i(<*-ahya) and plur. -ān (< *-ānām), even in the early stages of the most reduced Middle West dialect, Middle Persian (see CASES). Typologically, the function of the oblique, which derives from an original genitive/dative, was extended, as the morphologically marked form, to also, or exclusively, mark that of specific direct object. In several dialects this was later further disambiguated by adpositions (e.g., Sangsarī dir. obj. 3rd sing. near deictic nē-de, where is the oblique, and -de the specific direct object marker; see CASES).

Gender. Most notably, the distinction between masculine and feminine gender is preserved in many eastern as well as western dialects. Most conservative in terms of retention is Pashto, the nominal inflection of which also retains reflexes of the free accentual system of Indo-Iranian.

Iranian dialectology and history. In terms of retention, then, the two extremes are Eastern Iranian Pashto and Western Iranian Persian. And, contrary to an impression still widely advertized, modern Western Iranian dialects on the whole may be more conservative than Eastern Iranian. It may be noted that all evidence for the retention of the nominal oblique in Middle and Modern dialects, East as well as West, follows the merger of the earlier distinctions in Old Persian, rather than that in Avestan.

Some of the problems involved in the study of dialects include the criteria for groupings. They are often based on the selection of a few features. Nevertheless, even on the basis of a few forms, specialists are often able to eeconstruct the full pattern of the paradigm involvred. Another problem is that descriptions are often based on the information of only one or few informants, usually from the older generation, on the assumption that they preserve an older, and thus more original, stage of their dialect. Only more recently have generational differences and those in local subdivisions been addressed (e.g., Yarshater, 1969). In general, the systematic study of the socio-linguistic aspect of dialectology has been largely neglected. One such neglected study is social affiliation, which often involves diglossia, i.e., the use of different varieties or styles of a dialect or language, as well as bilingualism and multilingualism, characterized by a speaker using different languages in different social contexts, or language switch, and language loss, all contributing to linguistic diversity and groupings. Similarly, studies of the effects of linguistic substrates, adstrates, and superstrates is little developed. Interference between Iranian and non-Iranian, however, especially on the margins of the Iranian speaking areas, has attracted the continuous attention of scholars, such as Iranian and Armenian, Turkic, Arabic, Indo-Aryan, and Dravidian Brahui.

Overall, the available data, not only for the earlier stages, result in frustration. Nevertheless, the objective of clearly identifying the synchrony of a linguistic system as a functioning system at any one point in time, so as to establish systemic certainty prior to comparing items or parts thereof, is increasingly being accomplished.

The focus on selective features and feature clusters allows one to trace at least to some degree larger movements where historical evidence is lacking. In his study of the origins of Kurdish, MacKenzie (1961, pp. 68-86) suggests a path and diachrony of the moves of the Kurds and other Iranian groups into and within northwest Persia and the Fertile Crescent. Windfuhr (1975) suggests a symbiotic area of pre-Persians, -Kurds, and -Baluch in the northeast prior to their moving into their present areas, namely southwest, northwest, and southeast Peria, respectively. As another example, the isolated change of *r > š found in some lexical items in Sangsarī in the foothills some 300 km east of Tehran should be linked to a similar change found in Middle Eastern Iranian. The distinctive formation of the present indicative based on the old present participle in *-ant is found in the Māzandarānī dialects; in Sangsarī and neighboring dialects; in Harzandī (where it is reduced to the subjunctive stem), now in northern Azerbaijan, and in Zāzā, now in eastern Turkey (e.g., Sangesarī: ner-end-ī, Zāzā: niš-ann-ān “I sit down”). This reflects the retention of a feature that may go back to Old Iranian and suggests earlier contingency among these dialects. Both Eastern Iranian Pashto and Western Iranian Kurdish have subjunctive/counterfactual formations with prefix be- and affix -ā- (e.g., Pashto perfect conditional I and III: lwedəlay w-āy, lwedəlay ba w-āy “if it had fallen, (then) it would have fallen”; Kurd: bī-kawt-im-āyā, (bi)kawt-(i/ā) bam (āyā) “if I fell I had fallen”).

The continuing dialectical dynamics is shown by the fact that these dialect groups have considerably adjusted to the dialectal features of their hosts on all levels, which provides evidence for the relative ease of change. This includes syntactic features. For example, it was early recognized that in the Southwestern dialects dependent nouns and adjectives follow the head noun, but vice versa in the Northwestern dialects. The seeming Southwestern feature is found in much of Western Iranian, except for Tātī and Ṭālešī and related dialects. The latter also have postpositions as opposed to Perside prepositions. But fossilized personal pronouns derived from *haca- (from; e.g., čemen “mine"< *haca mana “from me”) suggest an earlier stage with prepositions, as opposed to postpositions in Turkic.

There is much evidence for ethnic and linguistic dynamics through the ages. To cite just one indigenous source, the multilingual and erudite medieval Kurd Bitlisi in his Šaraf-nāma, a history of the Kurds written in chancery Persian with poetry and poetic verve, suggests that the Kurds’ origin may well be multiple. Known and assumed moves, by force or voluntarily, on a large or small scale, most recently as an effect of the wars of Persia, Afghanistan, and Tajikistan, continue to turn the dialectal kaleidoscope and hologram.




Major overviews: Grundriss. Iranistik I. Linguistik, HO I/4, Leiden, 1958.

I. M. Oranskiĭ, Vvedenie v iranskuyu filologiyu (Introduction to Iranian philology), Moscow, 1960.

Idem, Iranskie jazyk, Moscow, 1963, tr. J. Blau as Les langues iraniennes, Paris, 1977; tr. K. Kešāvarz as Moqaddama-ye feqh al-lōḡa-ye īrānī, Tehran, 1358 Š./1979.

Idem, Die neuiranischen Sprachen der Sowjetunion, 2 vols., Paris and The Hague, 1975.

Idem, Iranskie jazyki v istoricheskom osvescheniĭ (The Iranian languages in historical view), Moscow, 1979.

V. S. Rastorgueva, ed., Opyt istoriko-tipologicheskogo issledovaniya iranskikh jazykov (Essays on the historical-typological study of Iranian languages), 2 vols., Moscow, 1975.

Idem, ed., Osnovy iranskogo jazykoznaniya yazyki, 4 vols., Moscow, 1979-87.

H. Reichelt, in Grundriss der indogermanischen Sprach-und Altertumskunde II, Berlin and Leipzig, 1927, pp. 1-84.

R. Schmitt, ed., Compendium Linguarum Iranicarum, Wies-baden, 1989.

Sebeok et al., eds., Current Trends in Linguistic, 6 vols., The Hague, 1963-70, V, pp. 450-77; VI, pp. 9-135.

Comprehensive etymologies. Air Wb. Horn, Etymologie. Hübschmann, Persische Studien. Iran-isches Personennamenbuch.

Dialectology: Ī. Afšār, “Ketāb-šenāsī-e fārsī-e čāp šoda dar Īrān dar bāra-ye gūyešhā-ye Īrānī,” in Études Irano-Aryennesoffertes à Gilbert Lazard, Paris, 1989, pp. 3-27.

Idem, “Ketāb-šenāsī-e zabānhā wa lahjahā-ye īrānī,” FIZ 3, 1334 Š./1955, pp. 78-93.

B. ʿAlawī, Čašmhā-yaš, 3rd ed. 1357 Š./1978.

F. C. Andreas, Iranische Dialekt-aufzeichnungen aus dem Nachlass von F.C. Andreas, ed. A. Christensen, Berlin, 1939.

Č.-ʿA. Aʿẓamī Sangsarī and G. L. Windfuhr, Vāža-nāma-ye sangsarī bā moqaddama-ī az dastūr-e ān, Tehran, 1351 Š./ 1972.

H. W. Bailey, “Persia. ii Language and Dialects,” in EI1III, pp. 1050-58.

W. Boeder, “Ergative Syntax and Morphology in Language Change. The South Caucasian Languages,” in F. Plank, ed., Ergativity. Towards A Theory of Grammatical Relations, London and New York, 1979, pp. 435-80.

G. Bossong, Empirische Universalien-forschnung.Differentielle Objektmarkierung in den neuiranischen Sprachen, Tübingen, 1985.

L. L. Cavalli-Sforza, “Easiest To Get, e.g., Genes, Peoples and Languages,” Scientific American, Nov. 1991, pp. 104-10.

A. Christensen, Contributions à la dialectologie iranienne, 2 vols., Copenhagen, 1930-35.

D. I. Edel’man, Osnovnye voprosy lingvisticheskoĭ geografii. Na materiale indo-iranskikhyazykov (Fundumental problems of linguistic geography. Based on data of the Indo-Iranian languages), Moscow, 1968.

Idem, Sravitel’naya grammatika vostochnoiranskih yazykov. Fonologiya (Comparative grammar of the East Iranian languages. Phonology), Moscow, 1986.

R. E. Emmerick, “Indo-Iranian languages. The Iranian Languages,” in The New Encyclopedia Britannica, Macropaedia IX, Chicago, 1976, pp. 450-57.

A. Erhart, Struktura indoíránských yazykov (The structure of the Indo-Iranian languages), Brno, 1980.

B. Farahvašī, Vāža-nāma-ye ḵūrī, Tehran, 1355 Š./1976.

Th. Gamkrelidze and V. V. Ivanov, “The Ancient Near East and the Indo-European Problem,” Soviet Studies in History 2/2, 1983, pp. 1-53.

W. Geiger, “Allgemeine Übersicht über die Dialekte und ihre Gruppierung,” in Grundriss I/2, pp. 412-23.

G. A. Grierson, Linguistic Survey of India. X Specimens of Languages of the Ēranian Family, Calcutta, 1921.

H. Haarmann, Die indirekte Erlebnisform als grammatische Kategorie.Eine eurasische Isoglosse, Wiesbaden, 1970.

H. Haider and R. Zwanziger, “Relatively Attributive. The ‘ezâfeδ- Construction from Old Iranian to Modern Persian,” in J. Fisiak, ed., Historical Syntax, Paris and The Hague, 1984, pp. 137-72.

W. B. Henning, “The Ancient Language of Azerbaijan,” TPS, 1954, pp. 157-77.

W. L. Heston, Selected Problems in Fifth to Tenth Century Iranian syntax, Ph.D. Dissertation, University of Pennsylvania, 1976.

W. Ivanov, “The Gabri Dialect,” Revista degli Studi Orientali 15, 1934, pp. 1-58; 16, 1935, pp. 31-97; 17, 1936, pp. 1-39.

K. J. Krahnke, Linguistic Re-lationships in Central Iran, Ph.D. Dissertation, University of Michigan, 1976.

G. Lazard, “Dialectologie de la langue persane d’après les textes de Xme et XIme siècles ap. J.-C.,” NDA Tabrīz 13, 1340 Š./1961, pp. 241-58.

Idem, La langue des plus anciens monuments de la prose persane, Paris, 1963.

Idem, “La dialectologie du judéo-persane,” Studies in Bibliography and Folklore 8, 1968, pp. 77-98.

Idem, “Pahlavi, pârsi, dari: Les langues de l’Iran d’après Ibn al-Muqaffaʿ,” in Iran and Islam, ed. C. E. Bosworth, 1971, pp. 361-91.

Idem, “The Rise of the New Persian language,” in Camb. Hist. Iran IV, pp. 595-632.

Idem, “Le morpheme en persan et les relations actancielles,” Bulletin de la Société de linguistique 77/1, 1982, pp. 177-207.

Idem, “Actance Variation and Categories of the Object,” in F. Plank, ed., Ergativity. Towards A Theory of Grammatical Relations, London and New York, 1984, pp. 269-92.

Idem, “L’inférentiel ou passé distancié en persan,” St. Ir. 14, 1985, pp. 27-42.

Idem, “Formes et fonctions du passif et de l’antipassif,” Actances 2, 1986, pp. 7-57.

Idem, “Échelles de transitivité,” in G. François-Geiger, ed., La Transitivité et ses correlats, 1987, pp. 109-19.

P. Lecoq, “La classement des languages irano-aryennes occi-dentales,” in C.-H de Fouchécour and Ph. Gignoux, eds., Études Irano-Aryennes offertes à Gilbert Lazard, Paris, 1989, pp. 247-64.

W. Lentz, “Die nordiranischen Elemente in der neupersischen Literatursprache bei Firdôsī,” ZII 4, 1926, pp. 251-316.

D. N. MacKenzie, “Gender in Kurdish,” BSO(A)S 16, 1954, pp. 529-41.

Idem, Kurdish Dialect Studies, 2 vols., London etc., 1961-62.

Idem, “The Origins of Kurdish,” TPS, 1961, pp. 68-86.

Y. Māhyār Nawwābī, Zabān-e konūnī-e Āḏarbāyjān, Tabrīz, 1333 Š./1954.

O. Mann and K. Hadank, Kurdish-Persische Forschungen, I, II, III. 1-2, IV. 1-2, Berlin, 1906, 1932.

A. Marʿašī, Vāža-nāma-ye gūyeš-e Gīlakī, Rašt, 1363 Š./1984.

C. P. Masica, Defining a Linguistic Area. South Asia, Chicago, 1976.

M. Moʿīn, “Zabān o lahjahā dar Borhān-e qāṭeʿ,” in Borhān-e qāṭeʿ, ed. M. Moʿīn, I, pp. XCVII-CX.

G. Morgenstierne, Report on A Linguistic Mission to Afghanistan, Oslo, 1926.

Idem, Indo-Iranian Frontier Languages, 2 vols., Oslo, 1929-38.

Idem, Report on a Linguistic Mission to North-Western India, Oslo, 1932.

Idem, “Bemerkungen zum Wort-Akzent in den Gathas und im Pashto,” MSS 42, 1983, pp. 167-75.

M. Mortażawī, Zabān-e dīrīn-e Āḏarbāyjān, Tehran, 1365 Š./1986.

Neue Methodologie in der Iranistik, Wiesbaden, 1974 (Lentz’s Festschrift).

J. Payne, “The Decay of Ergativity in the Pamir Languages,” Lingua 51, 1980, pp. 147-86.

Idem, “Iranian languages,” in B. Comrie, The Languages of the Soviet Union, Cambridge, 1981, pp. 158-79.

Idem, “Iranian Languages,” in ed. B. Comrie, The World’s Major Languages, London, 1987, pp. 514-22.

L. Pirejko, “On the Genesis of the Ergative Construction in Indo-Iranian,” in F. Plank, ed., Ergativity. Towards A Theory of Grammatical Relations, London and New York, 1979, pp. 481-88.

N. Rāst, “Fehrest-e maʾāḵeḏ-e zabānhā wa lahjahā-ye Īrānī,” FIZ 1, 1332 Š./1953, pp. 1-40.

V. S. Rastorgueva, Opyt sravitel’nogo izucheniya tadzhiksikh govorov (Comparative study of the Tajiki dialects), Moscow, 1964.

G. Redard, “Panorama linguistique de l’Iran,” Asiatische Studien 8, 1954, pp. 137-48.

Idem, “Projet d’un Atlas linguistique de l’Iran,” in Akten des 24. Inter-nationalen Orientalisten-kongresses, Wiesbaden, 1959, pp. 440-44.

Idem, “État des travaux et publication. Quelques cartes onomasiologiques,” in Atlas linguistique des parlers iraniens. Atlas de l’Afghanistan, University of Bern Institut für Sprachwissenschaft, Arbeitspapiere 3, Bern, 1974, pp. 7-19.

A. V. Rossi, “La varietà linguistica nell’Iran achemenide,” AIOUN 3, 1981, pp. 141-96.

M. M. Sakhokija, Posessiv’nost’, perekhodnost’ i ergativ’nost’. Tipologocheskoe sopostavlenie drevnepersidskikh, drevnearmjanskikh i drevne-gruzinskikh konstrukciĭ (Posessivity, transitivity, and ergativity. Typological comparison of the Old Persian, Old Armenian, and Old Georgian constructions), Tbilisi, 1985.

R. Schmitt, “Aryans,” in EIr II, pp. 684-87.

H. J. Seiler, Relativsatz, Apposition und Attribut, Wiesbaden, 1960.

V. S. Sokolova, Ocherki po fonetike iranskihkh yazykov (Studies on the phonetics of the Iranian languages), Moscow, 1953.

J. S. Sorūšīān, Farhang-e Behdīnān, ed. M. Sotūda, Tehran, 1335 Š./1954.

M. Sotūda, Farhang-e Semnānī. Amṯāl o eṣṭelāḥāt o ašʿār, Tehran, 1356 Š./1977.

G. Steiner, “The Intransative-Passival Conception of the Verb in Languages of the Ancient Near East,” in F. Plank, ed., Ergativity. Towards A Theory of Grammatical Relations, London and New York, 1979, pp. 185-216.

D. L. Stilo, “The Tati Language Group in the Sociolinguistic Context of Northwestern Iran and Transcaucasia,” Iranian Studies 14, 1981, pp. 137-87.

P. Tedesco, “Dialektologie.” G. L. Windfuhr, “Isoglosses. A Sketch on Persians and Parthians, Kurds and Medes,” in Monumentum H. S. Nyberg II, Acta Iranica 5, Tehran and Liège, 1975, pp. 457-72.

Idem, “The Verbal Category of Inference in Persian,” Monumentum George Morgenstierne II, Acta Iranica 22, Leiden, 1982, pp. 263-87.

Idem, “A Spatial Model for Tense, Aspect, and Mood,” Folia Linguistica 19, 1985, pp. 415-61.

Idem, “Convergence. Iranian Talyshi and Turkic Azari,” Select Papers from SALA-7. South Asian Languages Roundtable Conference, Bloomington, Ind., 1987, pp. 385-405.

E. Yarshater, A Grammar of Southern Tati Dialects, The Hague, 1969.

Idem, “Iran and Afghanistan,” in Th. A. Sebeok, ed., Current Trends in LinguisticsVI, The Hague, 1970, pp. 669-89.

Idem, “Zabānhā wa lahjahā-ye īrānī,” in Dehḵodā, I, Moqaddama, pp. 9-25.

V. A. Zhu-kovskiy, Materialy dlya izucheniya persid-skikh’ narechiĭ (Materials for the study of Persian dialects), 3 vols., St. Petersburg, 1888-1922.



Originally Published: December 15, 1995

Last Updated: November 22, 2011

This article is available in print.
Vol. VII, Fasc. 4, pp. 362-370