The non-Indo-European Hurrian and Urartian languages are related to one another, but modern research indicates that Urartian should not be considered as a direct continuation of Hurrian.

HISTORY AND DEVELOPMENT

Languages using cuneiform writing and Anatolian hieroglyphs.

Hattic.

Hattic studies began in 1922 with the work of the German Assyriologist Emil Forrer. In 1935, Hans G. Güterbock, a German-born Orientalist, published a large group of texts containing Hattic material and in so doing completed the publication of the Hattic texts stemming from the Winckler excavations (1905-12). Important studies on the subject have continued to appear since then.

Hittite.

Hittite tablets from places outside of the Hittite capital are rare; only stray examples have been found--e.g., in Tarsus, Alalakh, Ugarit, and Amarna. These findings attest to the growth of a great Hittite empire, especially between c. 1400 and c. 1190 BC. Old Hittite, the written embodiment of the earliest Indo-European language that has been discovered so far, is known from some tablets preserved in an "old ductus" type of handwriting that was typical of copies from the Old Kingdom period (c. 1700-1500 BC). The intermediary "Dark Age" between c. 1500 and c. 1400 BC is sometimes referred to as the period of the so-called Middle Hittite language. Most of the Old and Middle Hittite texts, however, are preserved in copies from the later empire period.

The archives of Bogazköy-Hattusa have been found in various places in the citadel, in the Great Temple complex, and in the "House on the Slope." Although the majority of the texts are concerned with religious subjects (oracle texts, hymns, prayers, myths, rituals, and festival texts), these archives also contain material of historical, political, administrative, literary, and legal character. The cuneiform adopted by the Hittite scribes is a variant of a writing system of Mesopotamian origin that closely resembles the ductus and shapes prevalent in tablets of the 17th century BC (layer VII) from Alalakh (modern Atsana in southeastern Turkey). It is possible that the cuneiform script might have been introduced as a result of the Hittites inducing Syrian scribes to transfer their activities to the Hittite capital during the early part of the Old Kingdom, shortly after 1650 BC. It has also been posited, with good reason, that the newly acquired script was first used to write Akkadian and was only later employed for Hittite as well. In addition to the genres enumerated above, the "scholarly literature" deserves to be mentioned. This consists of the material considered by the scribes to be essential for their training; it includes word lists, omens, and ritual prescriptions, all reflecting an encyclopaedic approach aimed at complete coverage of the subjects concerned. The Sumerian texts found in these archives belong to this class of literature. For treaties and correspondence with foreign powers, Akkadian was used as the diplomatic language of that period. Therefore, both Sumerian and Akkadian formed part of the curriculum of the qualified scribes, these languages belonging to the "eight languages" found in the Hittite archives. (see also Index: Sumerian language)

In actual fact, the first decipherer of Hittite was the Norwegian scholar J.A. Knudtzon, who pointed out in 1902 that the language of the so-called Arzawa letters (i.e., Hittite)--found in the Amarna archive--had an apparent affinity with Indo-European. Because the cuneiform script had already been deciphered, Knudtzon, and Bedrich Hrozný after him, were able to "read" their texts. Thus their discovery consisted more in the interpretation than in the actual decipherment of the written material. The first series of German excavations, lasting from 1905 to 1912, produced about 10,000 tablets. It was work on this corpus that familiarized Hrozný with the contents of these tablets and led him to his epoch-making discovery that Hittite was indeed Indo-European (1915).(See also WRITING: Cuneiform.)

Palaic.

The Indo-European character of Palaic was first advocated by Emil Forrer (1922). Part of the text material is preserved on tablets in "old ductus." The knowledge of the limited vocabulary leaves much to be desired, but parallels--especially in the inflection of the noun, the forms of the demonstrative, relative, and enclitic pronouns, and the verbal endings--vouch for a close relationship to Hittite and Luwian.

Luwian.

As in the case of Palaic, the pioneering work on Luwian written in cuneiform was done by Emil Forrer (1922). Following this work, new text materials were published in 1953, closely followed by both grammatical and vocabulary studies as well as a standard dictionary of Cuneiform Luwian (1959).

The Anatolian hieroglyphic system has a long history, with its logographic beginnings dating back to early Hittite stamp seals of the 18th and 17th centuries BC; the youngest texts seem to date from the last quarter of the 8th century BC. The geographical range of the inscriptions is great, stretching from Sipylus and Karabel in the extreme west to Alaca Hüyük and Bogazköy-Hattusa in the north, Malatya, Samsat, and Tell Ahmar (Til Barsib) in the east, and Hama and ar-Rastan in the south. During the "Dark Age" of the 16th and 15th centuries BC, the early writing grew into a fully developed writing system with logograms (word-signs), syllabic values, and auxiliary signs. During the New Empire, the script was already in use for a multitude of purposes (rock inscriptions, seals, and wooden tablets for everyday use in the temple and the army). Whether an example of the empire period such as the Aleppo inscription already reflects the Luwian language is a moot question but seems likely. It is certain that the later inscriptions of the Neo-Hittite states were in Luwian.

The first attempts to decipher Hieroglyphic Luwian, made by the British archaeologist Archibald H. Sayce, were fortunate in some fundamental details, but it was not until the 1930s that systematic and mutually stimulating research by scholars of several countries led to the establishment of a number of syllabic values for the characters as well as to a correct analysis of the sentence structure of the inscriptions. In his publication of the (bilingual) Hittite royal seals (in 1940, 1942), Hans G. Güterbock bridged the gap between the inscriptions of the empire period and the late Neo-Hittite states; the seals found in the French excavations at Ugarit (in northern Syria) served a similar purpose. The most important recent finding was the discovery in 1947 by Helmuth T. Bossert, a German archaeologist, of the Karatepe bilingual inscriptions, written in Phoenician and Hieroglyphic Luwian.

On many points the Luwian vocabulary is still an enigma. The unity between the various Luwian dialects and the close relationship of Luwian to the other members of the Anatolian subgroup, however, is secured by several linguistic parallels, especially in the singular inflection of the noun, the forms of certain pronouns, the verbal endings, and a number of lexical (vocabulary) correspondences.

Hurrian.

Hurrian texts have been found in Urkish (Mardin region, c. 2300 BC), Mari (on the middle Euphrates, 18th century BC), Amarna (Egypt, c. 1400 BC), Bogazköy-Hattusa (Empire period), and Ugarit (on the coastline of northern Syria, 14th century). Amarna yielded the most important Hurrian document, a political letter sent to Pharaoh Amenhotep III. From Mari came a small number of religious texts; from Bogazköy-Hattusa, literary and religious texts; and from Ugarit, vocabularies belonging to the more "scholarly literature" described above and Hurrian religious texts in Ugaritic alphabetic script. Hurrian personal names, found in texts from many sites (Bogazköy-Hattusa, Alalakh, Ugarit, and especially Nuzu), constitute a second linguistic source of major importance.

The research on Hurrian started in the 1890s with simultaneous contributions by several scholars. Subsequently, Bedrich Hrozný (1920) and Emil Forrer (1919, 1922) discovered the presence of Hurrian material in the Bogazköy-Hattusa archives.

Urartian.

Archibald H. Sayce was the first scholar to devote his attention to Urartian in the 1880s and 1890s and continued his activities until 1932. More important were the philological contributions of the German historian Carl F. Lehmann-Haupt between 1892 and 1935. The first reliable description of Urartian grammar was published by the German Orientalist Johannes Friedrich (1933).

Next to the Urartian texts in cuneiform writing, there also existed an indigenous hieroglyphic script that is still undeciphered and is too meagrely represented to warrant a serious attempt.

Languages using a derivative of the Phoenician or Greek alphabet.

Phrygian.

In a publication of new material from Gordium, the U.S. archaeologist Rodney S. Young cautiously suggested that the Old Phrygian alphabet may be dependent on a prototype in use on the North Syrian or Cilician coasts. The old idea that the Phrygian alphabet was dependent on a Greek one (and not vice versa) need not be abandoned in that case. Historically, such a derivation would present no problems, because the presence of Greek settlements in these areas during the second half of the 8th century BC is amply attested to by both Assyrian annals and late Greek historical sources as well as by archaeological findings. Internal evidence from the Phrygian alphabet, presented by the French linguist Michel Lejeune, serves as proof for some researchers that that alphabet derived from the Greek one.

Lydian.

Important results concerning Lydian were reached using a strictly combinatory method; i.e., passages were compared that expressed similar contents in a slightly different manner in order to obtain a better understanding of the language's structure. This stage of the research culminated in a conclusive article by the Italian Piero Meriggi on the Indo-European character of Lydian (1936). Subsequently, other scholars published evaluations of the Lydian data, a dictionary, and a grammar book. The study of Lydian is hampered by many lexicological uncertainties, but there is at least a growing consensus on matters of grammar leading to the common notion that Lydian belongs to the Anatolian subgroup of Indo-European. The final obstacle to this classification as Anatolian was removed in 1959 by the Italian Onofrio Carruba, who proved that Lydian, like the other members of the Anatolian branch, does not possess a separate feminine gender. Lydian shares common features with Hittite, Palaic, and Luwian and should therefore be acknowledged, it seems, as a fourth independent member of the Anatolian subgroup.

Carian.

Lycian.

Sidetic.

LINGUISTIC CHARACTERISTICS

Non-Indo-European languages.

Both the Hurrian and the Urartian languages differentiate between stems and suffixes, but there is again no sharp distinction between noun and verb. Many suffixes may be added onto one another in a row, but within the often prolonged suffix series a detailed order is rigidly observed. Among the suffixes added to the noun, several subgroups are distinguished; one group might be compared to the case endings of the Indo-European languages. One of the most characteristic phenomena of this group is the distinction between a subject case (the "nominative") and an "agentive" case. The agentive marks the actor or subject of a transitive verb when the object is expressed by its counterpart, the "nominative." The subject case is characterized by a lack of ending on the stem; it marks the subject in nominal sentences (sentences without verbs) and occurs with intransitive verbs and as the object of transitive verbs.

Phrygian.

The Anatolian subgroup of Indo-European.

Grammatical characteristics.

The Anatolian inflection of pronouns conforms to the traditional Indo-European pattern by being different from that of the nouns, but, at the same time, it shows some striking peculiarities. Typical Anatolian pronouns are: Hieroglyphic Luwian amu, equivalent to Lycian amu, emu, emu "I, me" (compare Hittite nominative uk, accusative ammuk); and Hittite nominative zik, accusative tuk "you," as compared to ti, tu in Palaic. Some of the languages have enclitic pronouns; i.e., pronouns pronounced as being part of the preceding word. A demonstrative pronoun aba- ("that," but in some member languages also "this") is present in Hittite, Palaic, Cuneiform and Hieroglyphic Luwian, Lycian (ebe-), and Lydian (ebad "here, there"), and an interrogative or relative pronoun kui- (compare Latin quis) is common to Hittite, Palaic, and Cuneiform Luwian. The corresponding terms for kui- in Hieroglyphic Luwian, Lycian, and Lydian also seem to be phonetic variants of the same original pronoun.

The Anatolian verbal system is simple: it has two moods (indicative and imperative) and two tenses (present and preterite). There are some traces--either to be classified as debris or as the nucleus for a future development--of an aorist -s fixed to the stem; e.g., kaness- "to acknowledge" (compare Greek gi-gno-sk-o); kalless- "to call" (compare Greek kaleo, aorist é-kale-s-a). (The aorist is a verb form denoting action without reference to its duration or completion.) A mediopassive "voice" is present in Hittite (es-a-ri "he is seated"; ki-tta-ri "he is lying"), Palaic, Luwian, and perhaps in Lydian. (The mediopassive expresses a type of reflexive meaning ["He washes himself"] or passive meaning ["He is being washed"].)

Reduplication (repetition) of the verbal stem occurs in the entire Anatolian subgroup. It adds an iterative or intensive nuance to the meaning, but it does not function in a system of tenses as in Greek. Very typical of the Anatolian subgroup are verbal suffixes like the causative -nu- (compare Hittite war- "to burn," warnu- "to kindle," harg- "to perish," harganu- "to ruin, to destroy"). In principle, these formations can be built on any verbal stem whose meaning permits such an addition. It should be stressed that in Hittite a normal expression for a "state" consists of a nominal sentence (that is, a sentence without a verb but with a noun, an adverbial expression, or a participle as predicate); sometimes, however, the verb es- "to be" is used as the carrier of modal or temporal nuances. The total absence of the Indo-European perfect (describing a "state" resulting from a recently concluded action) becomes very clear by the usage of the adverb nawi "not yet," which occurs with a present tense in Hittite (but which would employ a perfect tense, such as "has been," in English and other Indo-European tongues).

Very characteristic of the Anatolian subgroup is a strong preference for the linking together of particles and enclitic pronouns to form "chains" that are placed at the beginning of the sentence or clause. The first component of such a "chain" usually is a stressed part of the sentence or otherwise a sentence connective (like nu in Hittite, a in Luwian).

Phonological characteristics.

Advocates of the so-called laryngeal theory (first proposed by the Swiss linguist Ferdinand de Saussure in 1879) have found their postulate partly confirmed by Anatolian data. This theory maintains that the different forms of certain words in the various Indo-European subgroups can be satisfactorily explained only by assuming that all the known Indo-European languages have lost certain guttural sounds (laryngeals) that were originally present in the parent speech. In 1927 the Polish linguist Jerzy Kurylowicz and the French scholar Albert L.M. Cuny announced their discovery that in Hittite an h sound was preserved in positions in which a laryngeal would have formerly been (compare Hittite hant- "front" to English anti-; Hittite pahhur "fire" but English pyre). But the Anatolian evidence for the laryngeal theory is certainly not without problems, and the adherents of the theory consider that other laryngeals disappeared in Hittite as well. (see also Index: laryngeal consonant)

Lexical data.

The relationship with the other subgroups.

Indo-Iranian languages

The Indo-Iranian tongues have been used as both administrative and literary languages. Old Persian was the administrative language of the early Achaemenian dynasty dating from the 6th century BC; and an eastern Middle Indo-Aryan dialect was the language of the chancellery of the Mauryan emperor Ashoka in India in the mid-3rd century BC. As literary languages, the Indo-Iranian languages were used in the texts of some of the world's great religions: Indo-Aryan for Buddhism, Hinduism, and Jainism, and Iranian for Zoroastrian and Manichaean texts. The oldest Zoroastrian texts are in dialects included under the name Avestan. Commerce, conquest, and religion spread the influence of these languages. Indo-Aryan languages, for example, penetrated deep into Southeast Asia; names in Indonesia and other areas and Sanskrit texts in Cambodia reflect this influence.

The close relation between the Iranian and Indo-Aryan groups has never been doubted. They share characteristic features that set them apart as a subgroup of Indo-European. The long and short varieties of the Indo-European vowels e, o, and a, for example, appear as long and short a: Sanskrit manas- "mind, spirit," Avestan manah-, but Greek ménos "ardour, force." (In the following examples, a macron ({macron}) indicates a long vowel; a breve () indicates a short vowel. The spellings used in this article for Indo-Aryan and Iranian forms are traditional transliterations for the most part. In some cases, more accurate phonetic symbols are used. These can be found in the International Phonetic Alphabet.) In instances in which some Indo-European languages have an a sound, Indo-Iranian has i as a reflex of Indo-European sounds called laryngeals--e.g., Greek pater "father," Sanskrit pitr-, Avestan and Old Persian pitar-. After stems ending in long or short a, i, or u, an n occurs sometimes before the genitive (possessive) plural ending am (Avestan -am)--e.g., Sanskrit martyanam "of mortals, men" (from martya-); Avestan masyanam (from masya-); Old Persian martiyanam. (see also Index: Sanskrit language, Avestan language, Persian language)

In addition to several other similarities in their grammatical systems, Indo-Aryan and Iranian have vocabulary items in common--e.g., such religious terms as Sanskrit yajña-, Avestan yasna- "sacrifice"; and Sanskrit hotr-, Avestan zaotar- "a certain priest"; as well as names of divinities and mythological persons, such as Sanskrit mitra-, Avestan mi ra- "Mithra." Indeed, speakers of both language subgroups used the same word to refer to themselves as a people: Sanskrit arya-, Avestan airya-, Old Persian ariya- "Aryan." (see also Index: sound change)

The Indo-Aryan and Iranian language subgroups also differ from each other in a number of linguistic features, among them that Indo-Aryan has an i sound representing an Indo-European laryngeal sound not only in initial syllables but generally also in interior syllables; e.g., Sanskrit duhitr- "daughter" (cf. Greek thugáter). In Iranian, however, the sound is lost in this position; e.g., Avestan dugdar-, du{voiced velar fricative con.}dar-. Similarly, the word for "deep" is Sanskrit gabhira- (with i for i), but Avestan jafra-. Iranian also lost the accompanying aspiration (a puff of breath, written as h) that is retained in certain Indo-Aryan consonants; e.g., Sanskrit dha "set, make," bhr, "bear," gharma- "warm," but Avestan and Old Persian da, bar, and Avestan garma-. Further, Iranian changed stops such as p before consonants and r and v to spirants such as f: Sanskrit pra "forth," Avestan fra; Old Persian fra; Sanskrit putra- "son," Avestan pu ra-, Old Persian pussa- (ss represents a sound that is also transliterated as ç). In addition, h replaced s in Iranian except before non-nasal stops (produced by releasing the breath through the mouth) and after i, u, r, k; e.g., Avestan hapta- "seven," Sanskrit sapta-; Avestan haurva- "every, all, whole," Sanskrit sarva-. Iranian also has both xs and s sounds, resulting from different Indo-European k sounds followed by s-like sounds, but Indo-Aryan has only ks; e.g., Avestan xsayeiti "has power, is capable," saeiti "dwells," but Sanskrit ksayati, kseti. Iranian was also relatively conservative in retaining diphthongs that were changed to simple vowels in Indo-Aryan.

Iranian differs from Indo-Aryan in grammatical features as well. The dative singular of -a-stems ends in -ai in Iranian; e.g., Avestan masyai, Old Persian cartanaiy "to do" (an original dative singular form functioning as infinitive of the verb). In Sanskrit the ending is extended with a--martyay-a. Avestan also retains the archaic pronoun forms yus, yuzm "you" (nominative plural); in Indo-Aryan the -s- was replaced by y (yuyam) on the model of the 1st person plural--vayam "we" (Avestan vaem, Old Persian vayam). Finally, Iranian has a 3rd person pronoun di (accusative dim) that has no counterpart in Indo-Aryan but has one in Baltic.

The original location of the Indo-Iranian group was probably to the north of modern Afghanistan, in the present-day states of Tajikistan, Uzbekistan, Kyrgyzstan, Turkmenistan, and Kazakstan, where Iranian languages are still spoken. From there, some Iranians migrated to the south and west, the Indo-Aryans to the south and east. From geographical references in the earliest Indo-Aryan literary document, the Rigveda, it is clear that the earliest settlement of Indo-Aryans was in the northwest of the Indian subcontinent. Migration did not take place at once; there was doubtless a series of migrations. The date of entry of the Indo-Aryans into the subcontinent cannot be precisely determined, though the beginning of the 2nd millennium BC is plausible and generally accepted. (see also Index: Veda)

There is heated controversy concerning the precise linguistic position of the language of the Indo-Iranian family first attested in Middle Eastern cuneiform texts of c. 1450-1350 BC. Some borrowed words and proper names appearing in these Hittite-Hurrian documents have been interpreted as belonging either to Indo-Iranian, to an Indic subgroup of Indo-Iranian that had not yet fully split, or to Indo-Aryan proper. Complete scholarly agreement on this issue has not been reached.

The identification of the Harappan peoples of the Indus Valley, whose writing has not yet been satisfactorily deciphered, also awaits further research; with it may come a possible answer as to whether Indo-Aryans encountered these people or whether their civilization had passed by the time the Indo-Aryans arrived on the subcontinent. Whatever the answers to these problems may be, the reasons for the split of the Indo-Aryans and Iranians are not known.

In the following presentation regarding Indo-Aryan documents as evidence for linguistic history, it should be borne in mind that almost all dates are approximations.

THE INDO-ARYAN LANGUAGES

Languages of the group.

New Indo-Aryan is represented by such modern vernaculars as Hindi and Bengali, which began to emerge from about the 10th century AD. These too have earlier and later stages, culminating in the present-day languages.

New Indo-Aryan languages accounted for about 490,000,000 speakers in India, or approximately 74 percent of the population in the early 1980s. Considering the approximately 85,000,000 Bengali speakers in Bangladesh, approximately 63,000,000 speakers accounted for by Punjabi and Sindhi in Pakistan, and 11,000,000 Sinhalese (Sinhala) speakers in Sri Lanka (formerly Ceylon), the total number of New Indo-Aryan speakers is well over 650,000,000. According to the latest Indian census, there are 547 mother tongues of the Indo-Aryan group in use within the bounds of postpartition (1947) India. Some of these are dialects that are used by few speakers; others are official state languages having 30,000,000 or 50,000,000 speakers. The major groups of New Indo-Aryan languages are given in Table 4. Structurally and historically, Hindi and Urdu are one, although they are now official languages of different countries written in different alphabets. The term hindi (also hindvi) is known from as early as the 13th century. The term zaban-e-urdu "language of the imperial camp" came into use in about the 17th century. In the south, Urdu was used by Muslim conquerors of the 14th century.

Many of the languages in Table 4 are official state languages, the media of education up to the university level and of official transactions. Hindi, written in the Devanagari script, is the co-official language (with English) of the Republic of India and is used as a lingua franca throughout North India. It has varieties according to the mother tongue of the area; e.g., Bombay Hindi and Calcutta Hindi. Each of the major state languages has several other dialects in addition to the standard dialect adopted for official purposes. Including the various dialects down to the village level, it can be said that a chain of communication stretches across North India such that each dialect forms a link with each adjacent dialect. On the level of official languages this is not so: a Gujarati speaker will not readily understand colloquial Bengali.

Historical survey of the Indo-Aryan languages.

Old Indo-Aryan.

Epic Sanskrit is so called because it is represented principally in the two epics, Mahabharata and Ramayana. In the latter the term samskrta "formed, polished" is encountered, probably for the first time with reference to the language. The date of composition for the core of early Epic Sanskrit is considered to be in the centuries just preceding the Christian era.

Classical Sanskrit is the language of the major poetic works (kavya), drama (nataka), tales such as the Hitopadesha and Pañca-tantra, and technical treatises on grammar, philosophy, and ritual. It was used not only by the poet Kalidasa and his predecessors Bhasa, a dramatist, and Ashvaghosa, a Buddhist author, in the first centuries AD but was also continued long after Sanskrit was a commonly used mother tongue; indeed, Sanskrit is a language of learned treatises and commentaries to this day. It is also used as a lingua franca among pandits (Brahmin scholars) from different areas of India.

Linguistic developments can be traced from the early Vedic of the Rigveda through the later Samhitas on to the late Vedic of Brahmana prose and sutras, culminating in the language described by Panini, which is tantamount to Classical Sanskrit. For example, the nominative plural form ending in -asas (devasas "gods") was already less frequent than -as in the Rigveda and continued to lose ground later; in Brahmana, -as (e.g., devas) is the normal form. There are numerous other changes evident. For example, the instrumental singular form of -a-stems ends both in -a and -ena (a pronoun ending) in the Rigveda, with the latter form predominating; thus, virya "heroic might" appears once, and viryena occurs ten times (from virya- "heroic might, act"). In later Vedic -ena is the usual ending. All the early Vedic forms are expressly classed as belonging to the sacred language (chandas) by Panini.

The verb also shows chronological differences. For example, the 1st person plural ending -masi (e.g., bharamasi "we bear") predominates over -mas in Rigvedic but not in the Atharvaveda; -mas becomes the normal ending later. Early Vedic distinguishes between the aorist, imperfect, and perfect tenses. The aorist is commonly used to refer to an action that has recently taken place; the imperfect is a narrative tense referring to actions accomplished in the distant past. The perfect form of the verb originally denoted, as in Greek, a state reached; e.g., bi-bhay-a "is afraid" (root bhi). From earliest Vedic, however, this was not always the use of the perfect. Although the grammarian Panini distinguished between the three tenses noted (he said the perfect is used to denote an action beyond one's ken), the perfect and imperfect both came to be used as narrative tenses.

There are also future forms of Vedic, formed with suffixes (-isya and -sya) and used from earliest times. A future form, composed of an agent noun of the type kartr- "doer" and followed, except in the 3rd person, by forms of the verb as "be" (e.g., kartasmi [karta asmi] "I will do"), was recognized as in common use by Panini but is rare in early Vedic.

Early Vedic had a category that went out of use by the late Vedic period of Brahmanas--the injunctive, which was formally a form with secondary endings lacking the augment, a prefixed vowel. The injunctive could be used to denote a general truth. A general truth can also be signified by the subjunctive, which is characterized by the vowel a affixed to the present, aorist, or perfect stem. Later Vedic retained the injunctive only in negative commands of the type ma vadhis "do not slay." The subjunctive also diminished slowly until it was no longer used; for Panini the subjunctive belonged to sacred literature. The functions of the subjunctive were taken over by the form called optative (and the future form).

Noun forms incorporated into the verb system are numerous in early Vedic. Rigvedic has forms with affixes ya and tva functioning as future passive participles (gerundives); e.g., vac-ya- "to be said," kar-tva- "to be performed, done." The Atharvaveda has, additionally, forms with -(i)tavya (hims-itavya- "to be injured") and -aniya (upa-jiv-aniya- "to be subsisted upon"). By late Vedic, the type with tva had been eliminated; Panini recognized as normal the types karya-, kartavya-, karaniya- "to be done." In Indo-Aryan, from earliest Vedic down to New Indo-Aryan, forms called absolutives (or gerunds) are used to denote the previous of two or more actions performed (usually) by one agent: "having done . . . he did"; for example, piba nisadya "sit down (nisadya "having sat down") and drink." Rigvedic uses tvi, tva, tvaya, (t)ya to form absolutives, but these were later reduced to two: tva with a simple verb or one compounded with the negative particle, and ya with a verb compounded with a preverb (a preposition-like form).

Early Vedic also uses various case forms of action nouns in the capacity of infinitives; e.g., dative singular -tave (da-tave "to give"), genitive singular -tos (da-tos), both from a noun in -tu, which also supplies the accusative ending -tum (da-tum). There are other types in early Vedic, but the nouns in -tu are important; in late Vedic the accusative -tum and the genitive -tos (construed with ish or shak "be able, can") became the norm. According to Panini, forms in -tum and dative singular forms of action nouns are equivalent variants: bhok-tum gacchati/ bhojanaya gacchati "He is going out to eat."

That some forms fell into disuse in the course of Indo-Aryan is natural; the above represent both chronological and dialectal modifications. Such change was recognized by Indian grammarians; e.g., Patañjali, of the mid-2nd century BC, noted that perfect forms of the type ca-kr-a "you did, have done" (2nd person plural) were not in use at his time; instead, a nominal (adjective) form kr-ta-vant-as was used, consisting of the past passive participle kr-ta- and an adjectival suffix -vant. Indian grammarians also recognized the existence of different dialects. Panini noted forms used by northerners (udicya) and easterners (pracya), as well as various dialectal uses described by grammarians who preceded him.

Earlier documents also afford evidence for dialect variation; e.g., the early Vedic of the Rigveda is a dialect in which the Indo-European l sound was for the most part replaced by r--pra "fill," pur-na- "full." This change accords with Iranian; e.g., Avestan p rna "full." These forms contrast with Latin plenus and Gothic fulls, with l. Other dialects kept l and r distinct. There are also doublets that have both r and l in words with Indo-European r: rohita-/lohita- "red." The variant with l can be assumed to belong to an eastern dialect. This variance accords with Middle Indo-Aryan evidence and the fact that such l forms become more numerous in the tenth book (mandala) of the Rigveda, which is demonstrably more recent than the most ancient parts of the Rigveda and dates from a time when the Indo-Aryans had progressed farther east than their original location on the subcontinent. The development of retroflex l- and lh- sounds (produced by curling the tip of the tongue upward toward the hard palate) from the retroflex sounds of d (nila- "nest" from nida-) and dh when occurring between vowels is another feature characteristic of some dialects, including the major dialect of the Rigveda. (see also Index: Veda)

Classical Sanskrit represents a development of one or more such early Old Indo-Aryan dialects. At this state, the archaisms noted above have been eliminated. Moreover, the accentual system of Classical Sanskrit is not the same as that of Vedic, which had a system of pitches; vowels had low, high, or circumflex (first rising, then falling) pitch, and the particular vowel of a word that received high pitch could not be predicted. In Classical Sanskrit, on the other hand, the accent was probably predictable. If the next to the last vowel was long, it received the accent; if not, the vowel preceding it was accented. The Vedic system survived at least to the time of Panini, who described it fully and did not restrict it to sacred language. (see also Index: stress)

For all this simplification, Classical Sanskrit is considerably more complex than Middle Indo-Aryan. In addition to the vowels a, i, and u (in both long and short varieties), it has r and l used as vowels. Consonant clusters occur freely, except in word final position, and the system of sound modification conditioned by the context, called sandhi, is fully operative. Moreover, in its grammatical system Classical Sanskrit maintains the dual number, seven cases in addition to the vocative form (which marks the one addressed), and a complex set of alternations. For example, to the nominative singular form agni-s "fire," correspond the genitive singular agne-s "of fire" the nominative plural agnay-as "fires," and the instrumental plural agni-bhis "with fires," with differing vowels in the second syllable. There are also separate sets of nominal (noun) and pronominal (pronoun) endings. Some nouns and adjectives inflect as pronouns; e.g., ekasmai, dative singular masculine-neuter of eka- "one."

The verb system of Classical Sanskrit also maintains complex alternations. In the present tense of the type bhav-a-ti "becomes, is," the stem (bhav-a-) remains unchanged throughout the paradigm except for lengthening of the -a- to -a- before v and m. But other verbs have vowel alternation; e.g., as-mi "I am," s-mas "we are"; e-mi "I go," i-mas "we go"; juhomi "I pour," juhumas "we pour." A distinction is observed between active and mediopassive endings: jan-ay-a-ti "engenders" with the active ending -ti, but ja-ya-te "is born" with the mediopassive ending -te. (Mediopassive verb forms are used for the passive, reflexive, and other meanings.)

Classical Sanskrit also has a rich system of nominal and verbal derivatives. Compound words are of the following kinds: copulative (dvandva) compounds such as matapitarau "mother and father" (also elliptic pitarau "parents"); the type like tat-purusa- "his man," in which the first member is equivalent to a case other than nominative; the type like bahu-vrihi "much-rice," in which the object denoted is other than that of any of the members of the compound (bahur vrihir yasya "He who has much rice"); and adverbial compounds (avyayibhava) of the type upagni (upa-agni) "near the fire." In addition, there are derivatives with affixes -tara- and -tama, such as priya-tara- "very dear" and priya-tama- "most dear" from the adjective priya-. Pronouns have derivatives equivalent to case forms; e.g., tatra "there," yatra "where," and kutra "where?" are equivalent to locative forms such as tasmin, yasmin, and kasmin. These can also be used without a noun.

Among the derivative verbal systems are the causative and the desiderative ("desire to"); the former has an affix -ay- (gam-ay-a-ti "makes to go," kar-ay-a-ti "has do") or, after roots in -a, -pay- (stha-pay-a-ti "sets in place"). The desiderative is formed with -sa- and reduplication (repetition of a part of the root)--di-drk-sa-te "desires to see" (root drsh). The desiderative also has an agent noun in -u--di-drk-s-u "who wishes to see."

Middle Indo-Aryan.

The most archaic literary Prakrit is Pali, the language of the Buddhist canon (c. 5th century BC) and of the later stories and commentaries of Theravada Buddhism. Pali represents essentially a western Middle Indo-Aryan dialect, though there are sufficient easternisms in the canon to have led some scholars to the view that the canon as it exists today is a recast of an original in an eastern dialect. To the Buddhist literature also belongs the Gandhari Dhammapada, the only literary text written in a dialect of the northwest. The Niya documents, official documents written in Prakrit dating from the 3rd century AD, also belong to the northwest. The earliest inscriptional Middle Indo-Aryan is that of the Ashokan inscriptions (3rd century BC). These are more or less full translations from original edicts issued in the language of the east (from the capital Pataliputra in Magadha, modern Patna in Bihar) into the languages of the areas of Ashoka's kingdom. There are other Prakrit inscriptions up to the 4th century AD, and Sanskrit was not used inscriptionally until the first centuries AD. Literary Prakrits other than Pali were also used in independent works and in dramas along with Sanskrit. (see also Index: Pali language)

According to Prakrit grammarians, Maharastri ("From the Maharashtra Country") is the Prakrit par excellence. It is the language of kavyas (epic poems) such as the Ravanavaha (also called Setubandha) from no later than the 6th century AD. Maharastri is also the language of lyrics in Rajashekhara's Karpura-mañjari (c. 900), the only extant drama written completely in Prakrit, and of verses recited by women in the classical drama of Kalidasa and his successors, though not earlier. The literary dialect used for conversation among higher personages other than the king and his captains in the drama is Shauraseni, while Magadhi is used by lower personages.

The language of the early Jaina canon, the final version of which was made in the 5th or 6th century AD, is called Ardhamagadhi ("Half Magadhi"); Jaina also used another literary dialect, called Jaina Maharastri in non-canonical works. The oldest poetic work in this is Vimala Suri's Paumacariya (c. 3rd century). Of other Prakrit dialects mentioned by grammarians, Paishaci (or Bhuta-Bhasa, both meaning "Language of Demons") is noteworthy; it is said to be the language of the original Brhatkatha of Gunadhya, source of the Sanskrit book of stories Katha-saritsagara.

Buddhist works were also written using a language that has been called Buddhist Hybrid Sanskrit. Among these works is the Mahavastu, the core of which is thought to date from the 2nd century BC. This language is a Middle Indo-Aryan dialect of indeterminate origin, which steadily became more Sanskritized in prose sections of later works.

The most advanced stage of Middle Indo-Aryan, Apabhramsha, was also used as a literary language. That there was literary creation in Apabhramsha by the 6th century is clear from an inscription of King Dharasena II of Valabhi, in which the King praises his father as being adept in Sanskrit, Prakrit, and Apabhramsha composition. Moreover, in the fourth act of Kalidasa's drama Vikramorvashiya there are Apabhramsha verses. Because Kalidasa probably lived in the 3rd or 4th century, literary composition in Apabhramsha is earlier still, if these verses are legitimate. There is a great deal of later literature in Apabhramsha, for the most part Jaina works; e.g., Paumacariu of Svayambhu (8th-9th century), Harivamsha-purana of Puspadanta (10th century), Sanatkumara-cariu of Haribhadra (12th century). (see also Index: Apabhramsá language)

Middle Indo-Aryan is characterized generally by the reduction of the complexities seen in Old Indo-Aryan. The vowel system was reduced by the merger of r (and l) sounds with vowels and the change of the diphthongs ai and au to the vowel sounds e and o; e.g., Pali accha- "bear" (Sanskrit rksa-), ina- "debt" (Sanskrit rna-), uju- "straight" (Sanskrit rju-), pucchati "asks" (Sanskrit prcchati), metti- "friendship" (Sanskrit maitri-), orasa- "breast-born, legitimate" (Sanskrit aurasa-). Moreover, -aya- and -ava- commonly contracted to -e- and -o-; e.g., Pali jeti "conquers" (Sanskrit jayati), odhi- "limit" (Sanskrit avadhi-). Final consonants were deleted, with the exception of -m, which developed to an -m sound before which a vowel was shortened (Pali bhariyam "wife"; Sanskrit bharyam). Together with the trend toward replacing variable consonant stems by unchanging stems in -a-, this change had serious consequences for the grammar. Consonant stems steadily disappeared and were transformed to stems ending in a vowel; e.g., to Sanskrit sharad- "autumn," sarit- "stream," and sarpis- "butter" correspond the Pali forms sarada-, sarita, and sappi-. Consonant clusters were also modified in Middle Indo-Aryan; e.g., Pali khetta- "field" (corresponding to Sanskrit ksetra-), Pali dakkhina- "right, south" (Sanskrit daksina), aggi- "fire" (Sanskrit agni-), punna- "full" (Sanskrit purna), and tanha- "thrist" (Sanskrit trsna-). The shortening of vowels before modified consonant clusters led to the use of short e and o sounds, which were unknown in Old Indo-Aryan; e.g., Pali semha- "phlegm" (Sanskrit shlesman), ottha- "lip" (Sanskrit ostha-).

The above phenomena are not restricted to Pali; they are pan-Middle Indo-Aryan. Differences between Pali and Ashokan and other Prakrits include the retention of voiceless stops (i.e., p, t, k) between vowels in Pali and Ashokan dialects; other Middle Indo-Aryan dialects modify them. The extreme development appears in literary Maharastri, in which unaspirated stops (pronounced without an accompanying audible release, or pull of breath) other than retroflexes (t, d) and labials (p, b) were deleted, aspirated stops (pronounced with an audible puff of breath) were replaced by h, retroflexes (pronounced by curling the tongue upward toward the hard palate) became voiced, and labials were replaced by v; e.g., loa- "world" (Sanskrit loka-), loana- "eye" (Sanskrit locana-), saha- "branch" (Sanskrit shakha-), padhai "recites, reads" (Sanskrit pathati), and savaha- "curse" (Sanskrit shapatha-).

Essentially on the same level are the dialects of Jaina texts, but in these a y glide prescribed by grammarians occurs when a consonant is elided: vayana- "face" (Sanskrit vadana-); sayala- "whole" (Sanskrit sakala-). In Shauraseni, on the other hand, voiceless stops (e.g., p, t, k) between vowels are voiced (e.g., become b, d, g, respectively); e.g., ido "hence" (Sanskrit itah); tadha "thus" (Sanskrit tatha). Though Pali and Ashokan are at an earlier level of development with respect to these changes, they share with the rest of the Middle Indo-Aryan dialects the replacement of voiced aspirated sounds between vowels by h: lahu- "light, unimportant" from laghu-; dahati "gives" (Sanskrit dadhati). Similarly, they share the change of dy- to j: joti- "light, brilliance" (Pali jotati "shines," Sanskrit dyotate). Pali and Ashokan, however, retain a y sound, changed to j in most other Prakrits; e.g., the pronoun ya- (feminine ya-), as in Sanskrit, opposed to ja-.

The deletion of stop consonants noted above resulted in vowel sequences within words that were unknown to Old Indo-Aryan. Similarly, the extent of sandhi modification was restricted in Middle Indo-Aryan. The Middle Indo-Aryan vowels i and u do not change to y and v before dissimilar vowels in compounds; e.g., Maharastri rattiandhaa- "dark of night" (Sanskrit ratry-andhaka-). In addition, the first of two contiguous vowels in different words is subject to deletion; e.g., Pali manas'icchasi (from manasa icchasi) "you wish in your mind."

In its grammatical system, Middle Indo-Aryan also reduced complexities. The dual number no longer exists as a separate category; for Sanskrit dvabhyam "by two," Middle Indo-Aryan has dohi(m), with the ending -hi(m) equivalent to the instrumental plural -bhis of Old Indo-Aryan. Among other changes is the replacement of the dative case by the genitive except in particular usages; e.g., the use of forms corresponding to the Old Indo-Aryan dative to denote a purpose.

In Middle Indo-Aryan, nominal and pronominal forms are no longer strictly segregated; e.g., Ashokan vijitamhi "in the kingdom" (also vijite) has a pronominal ending equivalent to Sanskrit -smin.

In the verb system, the contrast between active (-ti) and mediopassive (-te) endings was obliterated. Further, the Old Indo-Aryan distinction between aorist, imperfect, and perfect forms was eliminated. With few exceptions, the sigmatic aorist (an aorist form with s) provides the only productive preterite of early Middle Indo-Aryan: Ashokan ni-kkhamisu "they set out" (Sanskrit nir-a-kramisur). In later Prakrits verbally inflected preterites were generally eliminated; in their place was used the past participle. For example, in Shauraseni devi uva-visa, maharao vi a-ado "Sit down, my queen, the king also has arrived," the past participle a-ado (Sanskrit a-gatah) agrees with maha-rao "king" (Sanskrit maha-rajah) in number and gender. If the verb is transitive, the participle agrees with the direct object, and the agent is denoted by an instrumental form: in Jaina Maharastri, tena vi savvam sittham "He has told everything," tena "by him" denotes the agent, and sittham "told" (Sanskrit shistam) agrees with the neuter singular form savvam (Sanskrit sarvam). When no object is denoted, the verb is in the neuter singular. Old Indo-Aryan used both the participial construction and the finite verb; thus to Prakrit so vi tena samam gao "He also went with him" could correspond Sanskrit so'pi tena saha gatah or so'pi tena sahagamat (saha agamat). The Middle Indo-Aryan development eliminated the latter.

Alternations of the Sanskrit type as-mi, s-mas were eliminated in Middle Indo-Aryan; the predominant type of present tense was formed from an unchanging vowel stem (Pali e-ti, e-nti "go[es]").

Nominal forms of the verb system are of the same types as Old Indo-Aryan; e.g., the Pali future passive participle katabba- (Sanskrit kartavya-) "to be done," Shauraseni karania; Ardhamagadhi, Jaina Maharastri, and Maharastri karanijja- "to be done." The infinitive is commonly formed on the present tense stem, not on the root, as in Old Indo-Aryan. Thus Pali pappotum is formed on the present pappoti; Sanskrit praptum is formed on the root prap, present tense prapnoti.

Middle Indo-Aryan shows evidence of dialectal differentiation. The earliest documents that allow one to determine roughly the dialect distribution are Ashoka's inscriptions. These represent three major dialect areas: east, as in the inscriptions of Jaugada, Dhauli, and Kalsi; west, in Girnar; and northwest, in Mansehra and Shahbazgarhi. Characteristic of the east dialect area is final -e, corresponding to -o in the west and -as in Sanskrit; in the east dialect area l also regularly corresponds to r of the west and of Sanskrit. Moreover, in the east dialect area there is a tendency to insert a vowel within consonant clusters, while in the west and northwest one of the consonants is assimilated to the other without an intervening vowel. For example, to Sanskrit rajñas "of the king" corresponds Girnar rañño, Shahbazgarhi raño, Jaugada lajine. Northwest stands apart in retaining three spirant sounds, sh, s, s, which merge to s elsewhere. Ashoka's eastern dialect, from the Magadha country, shows an s sound for Old Indo-Aryan sh, s, s, rather than the sh sound typical of literary Magadhi. Grammatical features also show dialectal variation; e.g., the Ashokan dative singular form is -aya in the western dialects (Girnar atthaya "for the purpose of") but -aye in the east (Kalsi, Dhauli atthaye).

As noted above, the most advanced development of Middle Indo-Aryan is seen in Apabhramsha. Sound changes that are typical of Apabhramsha include the replacement of the vowel sound a by u in final syllables; e.g., karahu "you do, make," corresponding to karaha (karadha) in other Prakrits. From stems in -aya- develop forms in -au and nasalized -au (nasalization is here indicated by a tilde): bhadarau "honored one, king" (Prakrit bhattarayo), hau "I" (Ashokan hakam). Nasalization also appears in environments in which earlier m occurred between vowels; e.g., gau "village" (from gama, Sanskrit grama). Numerous other sound changes are evident, among them the development of -s(s)- between vowels into h: taho "of him" (from Prakrit tassa, Sanskrit tasya); hohinti "will be" (compare Pali hossati). Apabhramsha contractions, such as -aya- changing to -a and -iya to -i, foreshadow New Indo-Aryan, in which the development was extended; e.g., Apabhramsha paniu "water" (Old Indo-Aryan paniyam), Gujarati pani, Hindi pani. (see also Index: Gujarati language)

In other points Apabhramsha also presaged New Indo-Aryan. The interest of Apabhramsha lies in the fact that contracted forms presage the New Indo-Aryan opposition of masculine, neuter, and feminine nouns; thus, Apabhramsha -au, -au, -i, Gujarati -o, -u, -i (gayo, gayu, gai "went"), Hindi -a, -i (gaya, gai). The case system of Apabhramsha is also at a more advanced level of disintegration than that of earlier Middle Indo-Aryan, with the instrumental and locative plurals being identical in form (-ahi or -ehi for -a-stems) and instrumental singular forms also being used as locatives.

In the Apabhramsha verb system, present tense stems in -a predominate. Apabhramsha verb endings differ from those of other Prakrits. Most interesting is the 3rd person plural type kara-hi "they do," which coexists with karanti. The form kara-hi, corresponding to the 3rd person singular kara-i "he does," is formed on the model of the pair kara-u (1st person singular, "I do") and kara-hu (1st person plural, "we do"). Here again Apabhramsha comes close to New Indo-Aryan. Moreover, Apabhramsha has some causative formations that do not occur elsewhere in Middle Indo-Aryan but are known from New Indo-Aryan--bham-ada-i "causes to turn," Gujarati bhamare che "causes to turn round," and pais-ara-i "causes to enter," Gujarati p sare che "causes to enter, to penetrate."

Also noteworthy are two syntactic usages that closely parallel those present in New Indo-Aryan. The present participle is used as a conditional; e.g., jai hau mi tena sahu tau karantu to kim asamahie sahu marantu "Even if I had performed (karantu) ascetic acts with him, would I have died without mental concentration?" in which the participles karantu and marantu have the value of conditionals. In Sanskrit the conditionals a-kar-isya-m and a-mar-isya-m are used; but in speaking Gujarati a person would say jo hu . . . karat . . . to marat, and Hindi would have the forms karta . . . marta. The Apabhramsha gerundive in -iv(v)a or -ev(v)a can be used as an infinitive; e.g., pi-evae lagga "began to drink." This is the Gujarati construction pi-va lagyo "began to drink," in which pi-va is an inflected form of pi-vu, that is, a verbal noun (infinitive) corresponding etymologically to the Apabhramsha gerundive.

Influences on Middle Indo-Aryan.

Middle Indo-Aryan shows similar evidence of the influence of linguistically more advanced vernaculars on literary compositions. The Prakrits of elegant literary compositions must have been artificial, different in many respects from the vernaculars current at the time, though reflecting languages that were current at some former time. The Old Indo-Aryan and Middle Indo-Aryan stages, then, present a picture of concurrent vernaculars with dialects and literary languages influenced by the vernaculars; it is impossible to compartmentalize the different stages as beginning and ending at any definite date.

The literary languages borrowed words and suffixes from earlier languages. There are Prakritisms (i.e., forms of earlier Prakrits) in Apabhramsha; e.g., the genitive singular ending -ssa instead of -ho and 2nd person plural verb forms terminating in -ha instead of -hu. All the literary Prakrits had recourse to Sanskrit as a source for borrowing words. Words that were incorporated into the Prakrits from Sanskrit with no change in form are called samskrta-sama "identical with Sanskrit" (or tat-sama "identical with that") and are contrasted with words termed samskrta-bhava (tad-bhava) "whose origin is in Sanskrit"--that is, words that the grammarians can derive from Sanskrit by using certain rules. Another class of words, called deshya (or deshi) "belonging to the area, country," includes items that the grammarians cannot derive easily from Sanskrit and that are supposed to have been in use in particular areas from early times.

Many or most of the deshya words are indeed derivable from Sanskrit, but some are of Dravidian origin; e.g., akka "sister" (Telugu akka), atta "father's sister" (Telugu atta), appa "father" (Telugu appa), ura "village" (Telugu uru), pulli "tiger" (Telugu puli). Borrowing from Dravidian occurred also at earlier times; the Dravidians originally occupied territory much farther north than they did in Middle Indo-Aryan times. The Rgveda has such words as kunda "pitcher, pot," which is doubtless of Dravidian origin (Tamil kutam "pot"). Such borrowings become more numerous in later Sanskrit. It is not always certain that borrowing proceeded from Dravidian to Indo-Aryan, however, because Dravidian languages freely borrowed from Indo-Aryan. Thus, some scholars claim that Sanskrit katu "sharp, pungent" is from Dravidian, but others claim that it is a Middle Indo-Aryan form deriving from an earlier *krt-u "cutting" (root krt). (An asterisk [*] preceding a form indicates that it is not attested but has been reconstructed as a hypothetical form.) Whatever the judgment on any individual word, it is clear that Indo-Aryan did borrow from Dravidian, and this phenomenon is important in considering a group of sounds that sets Indo-Aryan apart from the rest of Indo-European -- the retroflexes. Without doubt the influence of Dravidian is to be considered as contributing to the extension of these sounds beyond their limited occurrence in inherited Indo-European items such as nida "nest" (from *ni-sd-o), is-ta "desired" (from *is-to), and stir-na "spread out" (from *str-no). The Munda languages (or, more generally, the Austro-Asiatic languages) are also a source of some borrowing into Indo-Aryan; e.g., Sanskrit jambala "mud" (Santali jobo).

In the 8th century AD, the philosopher Kumarila mentioned not only Dravidian but also Persian and Greek as sources of foreign words. Such borrowing can be traced back to early times. In the 6th century BC Darius counted Gandhara as a province of his kingdom, and Alexander the Great penetrated into northern India in the 4th century BC. From Iranian come words such as that meaning "inscription, writing, script"; in the northwest inscriptions of Ashoka the word is dipi (Old Persian dipi) and Sanskrit has lipi, the form in other Ashokan versions and in Pali. Also from Persian is Sanskrit ksatrapa "satrap"--Old Persian xsassa-pavan-. Of Greek origin are such mathematical and astronomical terms as Sanskrit kendra "centre" (Greek kéntron), jamitra "diameter" (diámetron), and hora "hour" (hora). Yavana "foreigner," originally the Greek word for Ionian, is known from as early as the time of Panini. Later, Arabic words such as tashli "trigon" came into Sanskrit.

The modern Indo-Aryan stage.

Some of the modern Indo-Aryan languages have literary traditions reaching back centuries, with enough textual continuity to distinguish Old, Middle, and Modern Bengali, Gujarati, and so on. Bengali can trace its literature back to Old Bengali carya-padas, late Buddhist verses thought to date from the 10th century; Gujarati literature dates from the 12th century (Shalibhadra's Bharateshvara-bahubali-rasa) and to a period when the area of western Rajasthan and Gujarat are believed to have had a literary language in common, called Old Western Rajasthani. Jñaneshvara's commentary on the Bhagavadgita in Old Marathi dates from the 13th century and early Maithili from the 14th century (Jyotishvara's Varna-ratnakara), while Assamese literary work dates from the 14th and 15th centuries (Madhava Kandali's translation of the Ramayana, Shankaradeva's Vaisnavite works). Also of the 14th century are the Kashmiri poems of Lalla (Lallavakyani), and Nepali works have also been assigned to this epoch. The work of Jagannath Das in Old Oriya dates from the 15th century. (see also Index: Bengali language, Marathi language, Oriya language)

Amir Khosrow used the term hindvi in the 13th century, and he composed couplets that contained Hindi. In early times, however, other dialects were predominant in the midlands (Madhyadesha) as literary media, especially Braj Bhasa (e.g., Surdas' Sursagar, 16th century) and Awadhi (Ramcaritmanas of Tulsidas, 16th century). In the south, in Golconda (Andhra, near Hyderabad), Urdu poetry was seriously cultivated in the 17th century, and Urdu poets later came north to Delhi and Lucknow. Punjabi was used in Sikh works as early as the 16th century, and Sindhi was used in Sufi (Islamic) poetry of the 17th-19th centuries. In addition, there is evidence in late Middle Indo-Aryan works for the use of early New Indo-Aryan; e.g., provincial words and verses are cited. (see also Index: Punjabi language, Sindhi language)

The creation of linguistic states has reinforced the use of certain standard dialects for communication within a state in official transactions, teaching, and on the radio. In addition, attempts are being made to evolve standardized technical vocabularies in these languages. Dialectal diversity has not ceased, however, resulting in much bilingualism; for example, a native speaker of Braj Bhasa uses Hindi for communicating in large cities such as Delhi. (see also Index: Hindi language)

Moreover, the attempt to establish a single national language other than English continues. This search has its origin in national and Hindu movements of the 19th century down to the time of Mahatma Gandhi, who promoted the use of a simplified Hindi-Urdu, called Hindustani. The constitution of India in 1947 stressed the use of Hindi, providing for it to be the official national language after a period of 15 years during which English would continue in use. When the time came, however, Hindi could not be declared the sole national language; English remains a co-official language. Though Hindi can claim to be the lingua franca of a large population in North India, other languages such as Bengali have long and great literary traditions--including the work of Nobel Prize winner Rabindranath Tagore--and equal status as intellectual languages, so that resistance to the imposition of Hindi exists. This resistance is even stronger in Dravidian-speaking southern India. The use of English as an official language entails problems, however, because with the use of state languages for education, the level of English competence is declining. Another danger faced is the agitation for more separate linguistic states, threatening India with linguistic fragmentation hearkening back to earlier days.

Characteristics of the modern Indo-Aryan languages.

Literary languages tend to become somewhat removed from the usual standard colloquial. Literary, or High, Hindi, for example, tends to replace some of the Perso-Arabic vocabulary with Sanskritic items, whereas literary Urdu makes great use of Perso-Arabic words. The gap is formalized in Bengali, in which a distinction is made between the highly Sanskritic language Sadhu-Bhasa and the colloquial standard called Calit-Bhasa.

Phonology.

,

kru

Vowels in sequence contracted in early New Indo-Aryan; e.g., Old Indo-Aryan ashiti became Middle Indo-Aryan asii, Hindi and Punjabi ssi, and Bengali asi "80." Further, ai and au sounds changed to e and o, and au to u, while iu developed into i. The diphthongs ai and au were retained well into the New Indo-Aryan period and are still pronounced in some areas; e.g., Braj Bhasa kr u "I do," kr i "he does." Middle Indo-Aryan -d- and -dh developed into the flaps r and rh; e.g., Prakrit sadia "woman's garment," Kashmiri, Lahnda, Hindi, Gujarati, Bhojpuri, Bengali, Oriya sari "sari"; and Prakrit padh- "recite, read," Sindhi p rh-nu, Lahnda prh- n, Hindi, Punjabi prh-na, Gujarati prh-vu, Marathi prh-n "study."

Stress is not generally contrastive in New Indo-Aryan as it is, for example, in English (e.g., noun "éxport," verb "expórt"), though different areas have different rules for placing major emphasis on a given syllable. For example, in Hindi, in which vowel length is pertinent, gilá "swallowed" has major stress on the last syllable, gila "wet," on the first. In Gujarati, on the other hand, vowel length is not pertinent; the stress position depends on which vowels occur in contiguous syllables and on the structure of the syllables, whether open or closed; e.g., júno "old," but dukán "store." In Bengali each syllable of a word receives about equal stress.

The sounds that most clearly distinguish Indo-Aryan from the rest of Indo-European are the voiced aspirate stops (gh and the like, pronounced with an accompanying audible puff of breath) and the retroflexes (t and so on, pronounced by curling the tongue upward toward the hard palate). In the outlying New Indo-Aryan areas, however, the sound system is reduced. Sinhalese has no aspirated stops, Assamese has no retroflexes, and Kashmiri has no voiced aspirates. The geographic position of these languages doubtless contributed to these losses: Sinhalese coexists with Tamil, Assamese is surrounded by Tibeto-Burman languages, and Kashmiri is on the border of the Iranian area. (see also Index: Sinhalese language, Assamese language, Kashmiri language)

New Indo-Aryan shows evidence of early dialect distribution; this is discernible by considering sound changes proper to each group. The eastern group (Assamese, Bengali, Oriya) has three important changes. Long and short i and u merged; e.g., Assamese nila, Oriya nil{half-open back vowel} ({half-open back vowel} is similar to the o of "coffee" in some English dialects), Bengali nil "blue-black" but Sanskrit nila; Assamese dhuli, Bengali dhulo, Oriya dhuli "dust" but Hindi dhul and Sanskrit dhuli. The vowel sound a of Middle Indo-Aryan was replaced by {half-open back vowel} in Bengali and Oriya and {back open vowel} (similar to the o of "hot" in southern British English) in Assamese in initial position and open syllables; e.g., Bengali m{half-open back vowel}ron, Oriya m{half-open back vowel}r{half-open back vowel}n, Assamese m{back open vowel}r{back open vowel}n "death"; Sindhi, m rno "mortal, death," Sinhalese mr n, Gujarati, Marathi mr n (compare Sanskrit marana-). Moreover, in this group a vowel is affected by the quality of the vowel in a following syllable. For example, in Bengali ami kori "I do," the verb root has o followed by i in the next syllable, but tumi k{half-open back vowel}ro "you do" has an {half-open back vowel} sound; similarly, ami kini "I buy" but tumi keno. As a result of vowel assimilation also, Assamese has an {half-open back vowel} sound instead of {back open vowel} representing Middle Indo-Aryan a: Assamese x{half-open back vowel}hur, Bengali sosur "husband's father" (compare Hindi s sur, Prakrit sasura-, Sanskrit shvashura-).

Assamese and Bengali are set off from Oriya. In the former two, Middle Indo-Aryan d and dh merge medially to d (then r) with a subsequent development to r in Assamese; e.g., Oriya darhi, Bengali dari, Assamese dari "beard"; Hindi, Gujarati darhi, Prakrit dadhia. Assamese is also distinguished from Bengali by several developments, among them the merger of Assamese retroflex sounds with dental sounds; e.g., Assamese ut "camel" but Bengali ut, Oriya ot{half-open back vowel}, Sindhi uthu, Lahnda, Pahari utth, and so on. Assamese also has s for earlier c and ch sounds and a z sound for j and jh; e.g., Assamese kas "glass," Bengali kac; Assamese azi "today," Oriya aji, Bengali, Hindi aj. In addition, Assamese replaced an s sound initially by x and between vowels by h--x{half-open back vowel}hur.

Particular sound changes also characterize languages of the northwest. In this group, an older voiceless stop (e.g., t) became voiced (e.g., became d) after a nasal sound; in other areas, the voiceless stop is retained: Kashmiri dand, Punjabi d nd, Sindhi dndu "tooth" (the d in Sindhi is an imploded stop; see below) but Assamese, Bengali, Hindi, Gujarati, Marathi dãt, Sinhalese dt (Sanskrit danta-). Moreover, in the northwest group a voiced stop (e.g., d) preceded by a nasal was assimilated to the latter, resulting in two nasals, which were subsequently reduced to one in some areas; in the rest of New Indo-Aryan, the vowel preceding the nasal was nasalized. Thus, Kashmiri don "churning stick," Sindhi d nu "tribute," Punjabi dnn "fine," Lahnda dnn "force," Kumauni dan "roof" contrast with Assamese dãr "pole," Bengali dãr "oar," Hindi dãd "oppression, fine," and others; all forms derive from Old Indo-Aryan danda- "stick, staff, club, royal power, fine, punishment." (see also Index: Sinhalese language)

In the sequence of a short vowel followed by two consonants, Pahari differs from the rest of the northwest group and agrees with the rest of New Indo-Aryan. In the northwest this sequence either remained unchanged or the cluster was simplified without lengthening of the vowel; other languages generally simplified the cluster and lengthened the vowel: Punjabi bh tt, Sindhi bhtu, Lahnda bht, Kashmiri bat{central close vowel} "cooked rice, food" but Nepali, Kumauni, Hindi, Assamese, Bengali, Gujarati, Marathi bhat.

Dardic occupies a special position. The sibilant sounds did not all merge here. For example, Kashmiri, a Dardic tongue, has surah "16" with s rather than s, as in most other Indo-Aryan languages, and sat "7" with s. Further, voiced aspirated stops merged with unaspirated stops in Dardic; e.g., Kashmiri gur "horse" but Hindi ghora; Kashmiri d{half-open back vowel}d "milk" but Hindi dudh. (see also Index: Dardic languages)

One major feature distinguishing Sindhi from the rest of the northwest group is the development of a series of imploded stops (also called suction stops and recursive stops), for b, d, j, and g. Implosive stops also occur in the Sindhi vicinity; for example, Kacchi has imploded b. Another feature that distinguishes Sindhi from other northwest languages, including Kacchi, is the retention of the Middle Indo-Aryan final short vowels; e.g., Sindhi khi "eye" but Hindi ãkh (Middle Indo-Aryan akkhi-).

Punjabi is distinguished from other members of the northwest group by its tonal system, having low (`), mid ({macron}), and high () tones. Initial voiced aspirated stops of earlier Indo-Aryan appear in Punjabi as voiceless stops with low tone on the following vowel; e.g., Punjabi kòra but Hindi ghora; Punjabi tài "2 1/2" but Hindi dhai. Non-initially, a voiced aspirate became unaspirated and the preceding vowel received high tone; thus, Punjabi dd "milk" but Hindi dudh, and Punjabi láb "profit" but Hindi labh.

Gujarati, Marathi, and Konkani in the west and southwest differ from the languages of the midlands in that, as in the east, there is no contrast between long and short i and u vowels. The i of Gujarati and Marathi vis "20" is pronounced like the ee of English "teeth," the i of Gujarati iccha and Marathi iccha "wish" like the i of "pitch," but such a difference is not contrastive, as it is in Hindi (gila "wet": gila "swallowed"). Gujarati has certain features that, in turn, set it apart from the other languages of this group. In addition to e and o sounds, it has the open vowels , {half-open back vowel}; e.g., c{half-open back vowel}thu "fourth" (Middle Indo-Aryan cauttha), bs-vu "to sit" (Middle Indo-Aryan baisai "sits"). Moreover, Gujarati has murmured vowels, generally developed from vowels followed by h; e.g., k h che "says" (h represents murmuring of the vowel), Old Gujarati kahai chai. Marathi and Konkani have two series of affricate sounds; e.g., c (pronounced as the ch in English "chat"; the equivalent of c in some other languages) and c (pronounced as the ts of "rats").

There was clearly mutual influence of Indo-Aryan languages at an early time, together with movement of groups of speakers (compare the position of Pahari). Thus, while Punjabi s cc "true" is the expected form comparable to Middle Indo-Aryan sacca- (Old Indo-Aryan satya-), Hindi s c "true" does not represent the expected outcome. The item sc must come from the Punjabi area.

Grammar.

Over a large area of New Indo-Aryan the noun has only two cases--direct and oblique. A lack of distinction between direct and oblique cases in the plural is typical of several languages, including forms in Hindi, Gujarati, Marathi, and Bhojpuri. Direct forms are used independently, oblique forms before postpositions (words or word elements following a noun that function similarly to English prepositions) and other affixes; the combination of stem and postposition serves the function of inflected case forms of earlier Indo-Aryan. Thus, to denote an object (direct or indirect) Hindi uses the postposition ko, which occurs in direct object constructions normally only with nouns denoting animate beings; e.g., l rke-ko dekh-ta h "He sees the boy," lrke-ko mithai do "Give a sweet to the boy." Other postpositions are me "in," pr "on," se "from, with, by means of." A large group of postpositions are linked to the noun with the affix ka (oblique form ke, feminine ki), which also is used to form adjectives (possessives); e.g., l rke-ke sath gya "He went with the boy," lrke-ke pas h "The boy has it" (literally, "It is by the boy"). Many such postpositions represent old nominal (noun) forms. Other New Indo-Aryan languages have systems similar to that of Hindi, though the forms of the postpositions differ.

Though the nominal (noun) system of Punjabi is very close to that of Hindi, it has separate ablative (indicating separation and source) and locative (indicating place) forms in the singular and plural, respectively, for nouns such as kotha "house"; e.g., kothiõ "from the house," kothi "in the houses." Some languages have a fuller case system than that noted above; e.g., Bengali has a genitive singular ending, a genitive plural ending, and a locative case. Similarly, Kashmiri has nominative, dative, ablative, and agentive cases. Not all such case forms are inherited from Middle Indo-Aryan. In addition to case endings, these languages also use postpositions; e.g., Kashmiri garajas-andar "in the garage," with -andar after the dative ending -as.

Adjectives behave generally in the same way as nouns but have a syntactic restriction. In Hindi the possessive is in the oblique (non-nominative) form, as is the noun after which it occurs; but in the plural, only the noun has the oblique form. Further, the formation of comparatives and superlatives with derivative affixes has been eliminated. To a Sanskrit sentence such as ime amu-bhyah adhya-tarah "These (people) are richer than those," in which the comparative adhya-tara occurs construed with the ablative form, corresponds a Hindi sentence ye un-se mir h{nasal half-open front vowel}, in which no comparative affix is used--literally, "These are rich from (i.e., in comparison with) those." Comparable constructions with a postposition meaning "from" occur elsewhere in New Indo-Aryan.

The pronominal system of New Indo-Aryan formally resembles the Middle Indo-Aryan stage more than its noun system. For example, Gujarati hu "I," m{nasal half-open front vowel} "I" (agentive), me "we" (also agentive) are directly comparable to Apabhramsha hau, mai, amhai. The number distinctions of the Middle Indo-Aryan pronoun have been replaced, however, by distinctions of familiarity and politeness. For example, Hindi and Bengali have a three-way distinction--Hindi ap, Bengali apni "you" are polite or honorific forms; Hindi tum, Bengali tumi are informal forms; and Hindi tu, Bengali tui are used only for inferiors and small children. (Hindi and Bengali differ, however, in the plural forms of these.) In Gujarati, on the other hand, tu~ is a very familiar pronoun, whereas t me is used generally, covering the approximate domains of Hindi ap and tum; ap, if used, strikes the hearer as fawning. Marathi has a similar system. Southwestern languages also make a distinction in the 1st person plural between inclusive and exclusive, the exclusive excluding the person spoken to. In the form of the relative pronoun and the 3rd person pronoun, languages differ in the degree to which gender distinctions are made, thus contrasting with Old and Middle Indo-Aryan, in which these forms had three genders. For example, Marathi has masculine, feminine, and neuter for the relative pronoun, while Bengali has animate and inanimate.

New Indo-Aryan languages differ in the degree to which finite verb forms have been replaced by nominal (noun) forms. In Bengali a contrast is made between continuous or actual present (English "be . . . -ing") and non-continuous or habitual present; e.g., ami kaj kor-i "I work" (literally, "I do work"), with the ending -i, contrasts with ami kaj kor-ch-i "I am working," in which ch intervenes between the root and the ending. Hindi has a similar contrast but uses nominal forms; e.g., m{nasal half-open front vowel} kam kar-ta hu "I work," m{nasal half-open front vowel} kam k r rh-a hu "I am working." Both contain the finite form hu of the auxiliary; but kr-ta and rh-a are nominal forms, the latter the past of r h-"stay." Gujarati has both types, the present tense using finite verb forms, the imperfect employing nominal forms; e.g., hu kam kru chu "I work, am working" and hu kam kr-to h -to "I was working, used to work." Even in areas in which finite forms are not used in the present, they occur in the imperative forms and what may be called the subjunctive; e.g., Hindi tum kam kr-o "work," m{nasal half-open front vowel} ndr au "May I come in?"

The person-number system of the New Indo-Aryan verb accords with the use of pronouns. For example, the forms ja-o, kr-o in Gujarati tme kyã jao cho "Where are you going?" and su kro cho "What are you doing?" are historically plurals but are used with reference to one person addressed by the pronoun t me. Similarly, in Hindi, in which a person distinction is not made in the plural, ap k hã ja rhe h{nasal half-open front vowel}, ap kya kr r he h{nasal half-open front vowel}, equivalent in meaning to the Gujarati sentences, have the plural form rhe h{nasal half-open front vowel}. Bengali has completely given up any number distinction in verb forms: ami/amra kori "I/we do." In the 3rd person a distinction is made between ordinary and honorific: se (ordinary)/tini k{half-open back vowel}ren, plural tara/tãra k ren. Other languages (e.g., Hindi) also have honorific forms, for which the plural is used.

In the formation of the future there are again regional differences. Some retain the future in -s- (Gujarati hu kr-is, 3rd person e kr-s-e) or -h- (e.g., eastern dialects of Braj Bhasa, clih õ "I will go"). Characteristic of the Eastern languages and of Bihari (including Bhojpuri, Magahi, Maithili) is the suffix -b-; e.g., Bengali jabe "will go." All of these are finite forms. On the other hand, in Hindi and adjoining areas, the future is inflected for gender.

A similar contrast between the use of verbal and nominally inflected forms also appears in the past tense forms. The predominant pattern in New Indo-Aryan is that of Middle Indo-Aryan: forms are used that are etymologically participles.

The New Indo-Aryan languages retain the passive and causative forms. The causative is conservative in retaining both the affixes that appear in Middle Indo-Aryan and vowel alternation. The passive is also formed by affixation in some areas. But many languages also have a compound formation involving the verb ja "go" and an auxiliary (h{nasal half-open front vowel}); e.g., Hindi yahã hindi bol-i ja-t-i h{nasal half-open front vowel} "Hindi is spoken here."

There are other auxiliaries, which, like h{nasal half-open front vowel}, can occur with any verb in the language; e.g., the verb "can," Hindi s k-, Gujarati sk. A characteristic feature of New Indo-Aryan, however, is the use of certain verbs, variously called vector verbs or compound verbs, in restricted contexts and with particular semantics. For example, one can say mr g-ya "He died," bhul g-ya "He forgot," bol uth-a "He blurted out" in Hindi, using the verbs ja "go" (masculine singular past g -ya), uth "stand up." This phenomenon is pan-Indo-Aryan and still requires investigation.

The examples cited above also illustrate the normal word order in New Indo-Aryan languages: subject (including agential forms), object (with attributive adjectives preceding), verb (together with auxiliaries). Adverbials can precede the full sentence or occur after the subject, with slight differences in emphasis; e.g., Hindi m{nasal half-open front vowel} kl auga, or kl m{nasal half-open front vowel} auga "I will come tomorrow ( kl)." Relative clauses normally precede correlatives: Hindi jo admi kl tumhare ghr-me tha vo k{half-open back vowel}n h "Who (k{half-open back vowel}n) is the man (admi) who (jo) was in your house yesterday?" A notable exception to the normal final position for verbs occurs in Kashmiri, in which the verb usually occurs in second position after the subject; thus, to Hindi vo kha r ha h "he is eating" corresponds Kashmiri su chu khavan with the auxiliary chu after the subject.

Vocabulary.

n maz

msjid

m zhb.

svar

dal t

imart

mkan

mh l

k lm

d b

g rdn

sr

v kt

h fta

umr

r t

gr

m gr

The colloquial language used by any Hindu or Muslim communicating in Hindi-Urdu will contain a large number of such words. There have been efforts to polarize the two, and at times champions of Indo-Aryan have tried to replace Perso-Arabic vocabulary with Sanskritic words. The style that tends toward eliminating all but the most common Perso-Arabic words may be called High Hindi, written in the Devanagari script, as opposed to High Urdu, which retains Perso-Arabic of long standing, uses Persian and Arabic for learned vocabulary and is written in the Perso-Arabic script.

The influence of English as a source of borrowing still continues, and it is rare to hear a conversation on any technical subject among speakers of any Indian language in which English words are not liberally used. Among loanwords from English are names of conveyances such as Hindi rel-gari "railroad-train" and tksi "taxi"; profession names such as injinir "engineer," jj "judge," daktr "Western doctor," pulis "police"; and terms of educational administration such as kalj "college" and yunivrsiti "university." English words are susceptible to replacement in India by Sanskritic ones as are those of Perso-Arabic origin.

Of much lesser magnitude are New Indo-Aryan borrowings from other languages, among them Portuguese and Turkic. From the latter, the word urdu came to be used as the name of a language. From Portuguese come such Hindi words as nnnas "pineapple," pau "(Western style) bread," k miz "(Western) shirt," kmra "room," and girja "(Christian) church."

Writing systems.

The traditional alphabets are both over-explicit and not clear enough with regard to accurate representation of the spoken word. As systems in which a consonant symbol with no other accessory symbol accompanying it stands for the syllable consisting of the consonant followed by short a, they require previous knowledge of items for correct interpretation; Hindi krta is written ka-ra-ta in the Devanagari, and, to pronounce it properly, one must know that the word has only two syllables. Although Bengali has only the spirant sound s, the alphabet has symbols for sh, s, and s, as in Old Indo-Aryan; but verb forms such as kori and k ren are written ka-ri and ka-re-na, both with the same initial symbol. And, though syllabic r was lost as early as Middle Indo-Aryan, the scripts have a separate symbol for this. Script reform has been suggested; it has even been proposed that all Indo-Aryan languages adopt a Latin (roman) alphabet with diacritics, but chances for this are poor. (See also WRITING: Indian alphabets.) (Ge.Ca.) (see also Index: Bengali language)

THE IRANIAN LANGUAGES

Languages of the group.

Ancient (Old) Iranian.

The degree of mutual intelligibility that existed among the ancient Iranian languages is not known with certainty. The differences in the nature of the surviving sources have to be borne in mind. On the one hand, there is the religious poetry of Zoroaster in the Avestan language and, on the other, the official inscriptions of the Achaemenid rulers in Old Persian. Differences in the method of transmission present a further difficulty in the way of direct comparison. Nevertheless, it can safely be stated that the degree of mutual intelligibility must have been much greater between the ancient languages than between the Middle Iranian languages and that those languages geographically closer to each other probably were mutually understood better than those spoken in areas farther apart.

Avestan can hardly be said to be known beyond the ancient period, although only the earliest texts, the Gathas, are as old as the 6th century BC, and the later texts represent the language of several subsequent centuries. Old Persian, on the other hand, itself spanning the 6th to the 4th century BC, was continued more or less directly by the various forms of Middle Persian. Even in this case, however, although both Old and Middle Persian represent the language of the royal court, the considerable differences between them remain unexplained.

Middle Iranian.

Middle Persian and Parthian were doubtlessly similar enough to be mutually intelligible, but they differ so greatly from the eastern group of Middle Iranian languages that these must have appeared to be almost foreign languages. The languages of the eastern group, moreover, cannot have been themselves mutually intelligible. The main known languages of this group are Khwarezmian (Chorasmian), Sogdian, and Saka. Less well-known are Old Ossetic (Scytho-Sarmatian) and Bactrian, but from what is known it would seem likely that these languages were equally distinctive. There was probably more than one dialect of each of the languages of the eastern group, although there is certainty only in the case of Saka, for which at least two dialects are clearly attested. The main Saka dialect is known as Khotanese, but a small amount of material survives in a closely related dialect called Tumshuq, formerly known as Maralbashi. (see also Index: Saka language, Ossetic language)

A few words are known in all of these eastern Iranian languages from as early as the 2nd to the 4th century AD, but substantial evidence begins for Sogdian in the 4th century, for Saka probably no earlier than the 7th century (though that for Tumshuq may be a few centuries older), and for Khwarezmian not until the 12th century and later. The principal evidence for Bactrian belongs to the 2nd century. To the same period belong the Scytho-Sarmatian names of the earliest inscriptions.

All the eastern Iranian languages of the Middle Iranian period were spoken in Central Asia, with the exception of the language of the Scytho-Sarmatian inscriptions from what is now Ukraine, north of the Black Sea. More precisely, Bactrian was spoken in northern Afghanistan and in the adjacent parts of Central Asia. Khwarezmian was the language of Khwarezm, a historic region in present-day Turkmenistan and Uzbekistan but formerly of greater extent. Scholars believe that Sogdian was probably spoken over most of Central Asia, especially in eastern Uzbekistan, Tajikistan, and western Kyrgyzstan. There were also colonies of Sogdians in various cities along the trade routes to China; in fact, most Sogdian material comes from outside Sogdiana. The Saka dialects, Khotanese and Tumshuq, were spoken in Chinese Turkistan, modern Sinkiang; Tumshuq is the name of a small village in the extreme west of Sinkiang. Khotanese was spoken in Khotan near the modern city of Khotan (Chinese Ho-t'ien [Hotan]) on the southern route across the Takla Makan Desert and within about 100 miles (160 kilometres) to the north and to the east of Khotan, where manuscripts have been found, mainly at the sites of former shrines and monasteries.

Modern Iranian.

Only four of the many modern Iranian languages are the official languages of the state in which they are spoken. The chief of these is Persian (known in Persian as Farsi), the national language of Iran, which is spoken by about 27,000,000 people as a native language. A dialect of Persian known as Dari is recognized, moreover, as a second language in Afghanistan. The national language of Afghanistan is the East Iranian language known as Pashto, of which there are some 9,000,000 speakers, many living in Pakistan. Tajik is spoken by at least 7,000,000 people widely spread throughout Tajikistan and the rest of Central Asia and is readily intelligible to speakers of Persian, to which it is very closely related, although it is in some respects more archaic. (see also Index: Tadzhik language)

In addition to being the national language of Tajikistan, Tajik is important as the lingua franca of the Pamirs mountain range, a region where a remarkable variety of Iranian languages and dialects are spoken. Some 700,000 people speak Ossetic. Most of the Ossetes live in North Ossetia in Russia and South Ossetia in Georgia. Although spoken in the heart of the Caucasus Mountains, Ossetic is an East Iranian language not mutually intelligible with any other Iranian language.

Two other Iranian languages, Kurdish and Balochi (Baluchi), are spoken over a vast area, although they have not been officially accepted as the national language of an established state. Kurdish is spoken by more than 10,000,000 people living in Iran, Iraq, Turkey, Syria, and Transcaucasia. More than 5,000,000 people speak Balochi as their chief language; they are spread widely over parts of eastern Iran, Pakistan, Afghanistan, and Central Asia. In Iran, Balochi speakers live mainly in Baluchistan, a region in the southeast that now forms part of a province with Sistan. In Pakistan, Balochi speakers live mainly in the southwestern province of Balochistan; in Central Asia, they are found mainly around Mary (Merv) in southern Turkmenistan; and in Afghanistan, they are widely scattered, mainly over the southwestern portion of the country. There is a sizable Balochi colony in Oman, and many Balochi merchants have settled in the sheikhdoms of southern Arabia and along the east coast of Africa as far south as Kenya. Linguistically, Balochi and Kurdish are both West Iranian languages. Balochi is thus much more closely related to Kurdish than it is to its close neighbour Pashto. According to the most likely theory, the present eastern location of Balochi speakers is the result of migrations from the region of the Caspian Sea during the Middle Ages. (see also Index: Balochi language)

Dialects.

There are two main dialects of Ossetic: the eastern, known as Iron, and the western, known as Digor (Digoron). Of these, Digor is the more archaic, Iron words being often a syllable shorter than their Digor counterparts--e.g., Digor madä, Iron mad "mother." Iron is spoken by the majority of Ossetic speakers and is the basis of the literary language. Chosen in the 19th century for the translation of the Bible, it is still the official language today. Little is known of the other Ossetic dialects. A small amount of the Ossetic dialect of Tual in the south, which differs little from Iron, was published in Georgian script at the beginning of the 19th century.

Yaghnabi is still spoken by a small number of people southeast of Samarkand, Uzbekistan. It has two main dialects, eastern and western, which differ only slightly. The characteristic difference is between a western t sound and an eastern s sound from an older sound (as th in English "thin")--e.g., western met, eastern mes "day," beside Sogdian me (Christian Sogdian my).

Dialects of the Shughni group are spoken in the Pamirs. Closely related to this group is Yazgulami. A period of a Yazgulami-Shughni common language (protolanguage) has been postulated by some scholars, after which it separated first into Yazgulami and Common Shughni; and then Common Shughni gradually divided into Sarikoli, Oroshori-Bartangi, Roshani-Khufi, and Bajuvi-Shughni. Sarikoli, the easternmost of these dialects, is spoken in northwestern China.

Speakers of Wakhi number 10,000 or so in the region of the upper Pyandzh (Panj) River. Vakhan (Wakhan), the Persian name for the region in which Wakhi is spoken, is based on the local name Wux, a Wakhi development of *Waxsu, the old name of the Oxus River (modern Amu Darya). (An asterisk denotes a hypothetical, unattested, reconstructed form or word.) The Wakhi language is remarkably distinct from its neighbours and has many archaic features.

Around the bend of the Amu Darya and in the valley of the Varduj River to the southeast, a few people speak dialects of the Sanglechi-Ishkashmi group. This group is clearly distinguished from its neighbours but is closely related to the other languages of the Pamirs.

Some 6,000 people speak dialects of the Yidgha-Munji group. Monjan is a very remote valley located in northern Afghanistan, and it is separated by a mountain pass from the Sanglechi-speaking region. Yidgha is spoken in the valley of the Lutkho River and in the nearby city of Chitral, a region now in Pakistan. Yidgha-Munji is most closely related to Pashto.

The existence of two dialectal groups within Pashto has long been known. Thus, the word Pashto represents a southwestern dialect form (pasto), in contrast to a northeastern (paxto). According to one hypothesis, Pashto literature, which exists certainly from the 17th cent