Present-day geographical distribution of the major Indo-Aryan language groups.Romani,Domari,Kholosi,Luwati, andLomavren are outside the scope of the map.
The Indo-Aryan family as a whole is thought to represent adialect continuum, where languages are often transitional towards neighbouring varieties.[15] Because of this, the division into languages vs. dialects is in many cases somewhat arbitrary. The classification of the Indo-Aryan languages is controversial, with many transitional areas that are assigned to different branches depending on classification.[16] There are concerns that atree model is insufficient for explaining the development of New Indo-Aryan, with some scholars suggesting thewave model.[17]
The following table of proposals is expanded fromMasica (1991) (from Hoernlé to Turner), and also includes subsequent classification proposals. The table lists only some modern Indo-Aryan languages.
Anton I. Kogan, in 2016, conducted alexicostatistical study of the New Indo-Aryan languages based on a 100-wordSwadesh list, using techniques developed by the glottochronologist and comparative linguistSergei Starostin.[17] That grouping system is notable for Kogan's exclusion of Dardic from Indo-Aryan on the basis of his previous studies showing low lexical similarity to Indo-Aryan (43.5%) and negligible difference with similarity to Iranian (39.3%).[20] He also calculated Sinhala–Dhivehi to be the most divergent Indo-Aryan branch. Nevertheless, the modern consensus of Indo-Aryan linguists tends towards the inclusion of Dardic based on morphological and grammatical features.[citation needed]
TheInner–Outer hypothesis argues for a core and periphery of Indo-Aryan languages, with Outer Indo-Aryan (generally including Eastern and Southern Indo-Aryan, and sometimes Northwestern Indo-Aryan,Dardic andPahari) representing an older stratum of Old Indo-Aryan that has been mixed to varying degrees with the newer stratum that is Inner Indo-Aryan. It is a contentious proposal with a long history, with varying degrees of claimed phonological and morphological evidence. Since its proposal byRudolf Hoernlé in 1880 and refinement byGeorge Grierson it has undergone numerous revisions and a great deal of debate, with the most recent iteration byFranklin Southworth andClaus Peter Zoller based on robust linguistic evidence (particularly an Outer past tense in-l-). Some of the theory's sceptics includeSuniti Kumar Chatterji andColin P. Masica.[citation needed]
TheDardic languages (also Dardu or Pisaca) are a group of Indo-Aryan languages largely spoken in the northwestern extremities of the Indian subcontinent. Dardic was first formulated byGeorge Abraham Grierson in hisLinguistic Survey of India but he did not consider it to be a subfamily of Indo-Aryan. The Dardic group as a genetic grouping (rather than areal) has been scrutinised and questioned to a degree by recent scholarship: Southworth, for example, says "the viability of Dardic as a genuine subgroup of Indo-Aryan is doubtful" and "the similarities among [Dardic languages] may result from subsequent convergence".[21]: 149
The Dardic languages are thought to be transitional with Punjabi and Pahari (e.g. Zoller describes Kashmiri as "an interlink between Dardic and West Pahāṛī"),[22]: 83 as well as non-Indo-Aryan Nuristani; and are renowned for their relatively conservative features in the context ofProto-Indo-Aryan.
Northwestern Indo-Aryan languages are spoken in the northwestern region of India and eastern region of Pakistan.Punjabi is spoken predominantly in thePunjab region and is the official language ofthe northern Indian state of Punjab, in addition to being the most widely-spoken language in Pakistan.Sindhi and its variants are spoken natively in the Pakistani province ofSindh and neighbouring regions. Northwestern languages are ultimately thought to be descended fromShauraseni Prakrit, with influence fromPersian andArabic.[23]
Western Indo-Aryan languages are spoken in central and western India, in states such asMadhya Pradesh andRajasthan, in addition to contiguous regions in Pakistan. Gujarati is the official language ofGujarat, and is spoken by over 50 million people. In Europe, variousRomani languages are spoken by theRomani people, an itinerant community who historically migrated from India. The Western Indo-Aryan languages are thought to have diverged from their northwestern counterparts, although they have a common antecedent inShauraseni Prakrit.
TheEastern Indo-Aryan languages, also known asMagadhan languages, are spoken throughout the eastern subcontinent, includingOdisha andBihar, alongside other regions surrounding the northwestern Himalayan corridor.Bengali is the seventh most-spoken language in the world, and has a strong literary tradition; thenational anthems ofIndia andBangladesh are written in Bengali.Assamese andOdia are the official languages ofAssam andOdisha, respectively. The Eastern Indo-Aryan languages descend from MagadhanApabhraṃśa[24] and ultimately fromMagadhi Prakrit.[25][26][24] Eastern Indo-Aryan languages display many morphosyntactic features similar to those ofMunda languages, while western Indo-Aryan languages do not. It is suggested that "proto-Munda" languages may have once dominated the easternIndo-Gangetic Plain, and were then absorbed by Indo-Aryan languages at an early date as Indo-Aryan spread east.[27][28]
Marathi-Konkani languages are ultimately descended fromMaharashtri Prakrit, whereas Insular Indo-Aryan languages are descended fromElu Prakrit and possess several characteristics that markedly distinguish them from most of their mainland Indo-Aryan counterparts. Insular Indo-Aryan languages (ofSri Lanka andMaldives) started developing independently and diverging from the continental Indo-Aryan languages from around 5th century BCE.[17]
Some theonyms, proper names, and other terminology of the LateBronze AgeMitanni civilisation ofUpper Mesopotamia exhibit an Indo-Aryan superstrate. While what few written records left by the Mittani are either inHurrian (which appears to have been the predominant language of their kingdom) orAkkadian (the maindiplomatic language of the Late Bronze Age Near East), these apparently Indo-Aryan names suggest that an Indo-Aryan elite imposed itself over theHurrians in the course of theIndo-Aryan expansion. If these traces are Indo-Aryan, they would be the earliest known direct evidence of Indo-Aryan, and would increase the precision in dating the split between the Indo-Aryan and Iranian languages (as the texts in which the apparent Indicisms occur can be dated with some accuracy).
In a treaty between theHittites and the Mitanni, the deitiesMitra,Varuna,Indra, and theAshvins (Nasatya) are invoked.Kikkuli's horse training text includes technical terms such asaika (cf. Sanskriteka, "one"),tera (tri, "three"),panza (panca, "five"),satta (sapta, seven),na (nava, "nine"),vartana (vartana, "turn", round in the horse race). The numeralaika "one" is of particular importance because it places the superstrate in the vicinity of Indo-Aryan proper as opposed to Indo-Iranian in general or early Iranian (which hasaiva).[32] Another text hasbabru (babhru, "brown"),parita (palita, "grey"), andpinkara (pingala, "red"). Their chief festival was the celebration of thesolstice (vishuva) which was common in most cultures in the ancient world. The Mitanni warriors were calledmarya, the term for "warrior" inSanskrit as well; notemišta-nnu (=miẓḍha, ≈ Sanskritmīḍha) "payment (for catching a fugitive)" (M. Mayrhofer,Etymologisches Wörterbuch des Altindoarischen, Heidelberg, 1986–2000; Vol. II:358).
Sanskritic interpretations of Mitanni royal names renderArtashumara (artaššumara) asṚtasmara "who thinks ofṚta" (Mayrhofer II 780), Biridashva (biridašṷa, biriiašṷa) asPrītāśva "whose horse is dear" (Mayrhofer II 182), Priyamazda (priiamazda) asPriyamedha "whose wisdom is dear" (Mayrhofer II 189, II378), Citrarata asCitraratha "whose chariot is shining" (Mayrhofer I 553), Indaruda/Endaruta asIndrota "helped byIndra" (Mayrhofer I 134), Shativaza (šattiṷaza) asSātivāja "winning the race price" (Mayrhofer II 540, 696), Šubandhu asSubandhu "having good relatives" (a name inPalestine, Mayrhofer II 209, 735), Tushratta (tṷišeratta, tušratta, etc.) as *tṷaiašaratha, VedicTvastar "whose chariot is vehement" (Mayrhofer, Etym. Wb., I 686, I 736).
While Old Indo-Aryan is the earliest stage of the Indo-Aryan branch, from which all known languages of the later stages Middle and New Indo-Aryan are derived, some documented Middle Indo-Aryan variants cannot fully be derived from the documented form of Old Indo-Aryan (on which Vedic and Classical Sanskrit are based), but betray features that must go back to other undocumented dialects of Old Indo-Aryan.[34]
From Vedic Sanskrit, "Sanskrit" (literally 'put together, perfected, elaborated') developed as the prestige language of culture, science and religion, as well as the court, theatre, etc. Sanskrit of the later Vedic texts is comparable toClassical Sanskrit, but is largelymutually unintelligible with Vedic Sanskrit.[35]
Outside the learned sphere of Sanskrit, vernacular dialects (Prakrits) continued to evolve. The oldest attested Prakrits are theBuddhist andJain canonical languagesPali andArdhamagadhi Prakrit, respectively. Inscriptions inAshokan Prakrit were also part of this early Middle Indo-Aryan stage.
By medieval times, the Prakrits had diversified into variousMiddle Indo-Aryan languages.Apabhraṃśa is the conventional cover term for transitional dialects connecting late Middle Indo-Aryan with early Modern Indo-Aryan, spanning roughly the 6th to 13th centuries. Some of these dialects showed considerable literary production; theŚravakacāra of Devasena (dated to the 930s) is now considered to be the first Hindi book.
The next major milestone occurred with theMuslim conquests in the Indian subcontinent in the 13th–16th centuries. Under the flourishingTurco-MongolMughal Empire,Persian became very influential as the language of prestige of the Islamic courts due to adoption of the foreign language by the Mughal emperors.
In theCentral Zone Hindi-speaking areas, for a long time theprestige dialect wasBraj Bhasha, but this was replaced in the 13th century byDehlavi-basedHindustani. Hindustani was strongly influenced byPersian, with these and later Sanskrit influence leading to the emergence of Modern Standard Hindi and Modern StandardUrdu asregisters of the Hindustani language.[36][37] This state of affairs continued until the division of the British Indian Empire in 1947, when Hindi became the official language in India andUrdu became official in Pakistan. Despite the different script the fundamental grammar remains identical, the difference is moresociolinguistic than purely linguistic.[38][39][40] Today it is widely understood/spoken as a second or third language throughout South Asia[41] and one of the most widely known languages in the world in terms of number of speakers.
Domari is an Indo-Aryan language spoken by olderDom people scattered across the Middle East. The language is reported to be spoken as far north asAzerbaijan and as far south as central Sudan.[42]: 1 Based on the systematicity of sound changes, linguists have concluded that the ethnonymsDomari andRomani derive from the Indo-Aryan wordḍom.[43]
Parya is spoken inTajikistan andUzbekistan by the descendants of migrants from the Indian subcontinent. The language retains many features similar to Punjabi and the Western Hindi dialects, while also bearing some influence from Tajik Persian.[45]
The Romani language is usually included in the Western Indo-Aryan languages.[46] Romani varieties, which are mainly spoken throughout Europe, are noted for their relatively conservative nature; maintaining the Middle Indo-Aryan present-tense person concord markers, alongside consonantal endings for nominal case. Indeed, these features are no longer evident in most other modern Central Indo-Aryan languages. Moreover, Romani shares an innovative pattern of past-tense person, which corresponds to Dardic languages, such as Kashmiri and Shina. This is believed to be further indication that proto-Romani speakers were originally situated in central regions of the subcontinent, before migrating to northwestern regions. However, there are no known historical sources regarding the development of the Romani language specifically within India.
Research conducted by nineteenth-century scholars Pott (1845) and Miklosich (1882–1888) demonstrated that the Romani language is most aptly designated as a New Indo-Aryan language (NIA), as opposed to Middle Indo-Aryan (MIA); establishing that proto-Romani speakers could not have left India significantly earlier than AD 1000.
The principal argument favouring a migration during or after the transition period to NIA is the loss of the old system of nominal case, coupled with its reduction to a two-way nominative-oblique case system. A secondary argument concerns the system of gender differentiation, due to the fact that Romani has only two genders (masculine and feminine). Middle Indo-Aryan languages (named MIA) generally employed three genders (masculine, feminine and neuter), and some modern Indo-Aryan languages retain this aspect today.
It is suggested that loss of the neuter gender did not occur until the transition to NIA. During this process, most of the neuter nouns became masculine, while several became feminine. For example, the neuteraggi "fire" in Prakrit morphed into the feminineāg in Hindi, andjag in Romani. The parallels in grammatical gender evolution between Romani and other NIA languages have additionally been cited as indications that the forerunner of Romani remained on the Indian subcontinent until a later period, possibly as late as the tenth century.
Kholosi,Jadgali, andLuwati represent offshoots of the Sindhic subfamily of Indo-Aryan that have established themselves in thePersian Gulf region, perhaps through sea-based migrations. These are of a later origin than the Rom and Dom migrations which represent a different part of Indo-Aryan as well.
The use by theBritish East India Company of indentured labourers led to the transplanting of Indo-Aryan languages around the world, leading to locally influenced lects that diverged from the source language, such asFiji Hindi andCaribbean Hindustani.
The normative system of New Indo-Aryan stops consists of fiveplaces of articulation:labial,dental, "retroflex",palatal, andvelar, which is the same as that of Sanskrit. The "retroflex" position may involve retroflexion, or curling the tongue to make the contact with the underside of the tip, or merely retraction. The point of contact may bealveolar orpostalveolar, and the distinctive quality may arise more from the shaping than from the position of the tongue. Palatal stops haveaffricated release and are traditionally included as involving a distinctive tongue position (blade in contact with hard palate). Widely transcribed as[tʃ],Masica (1991:94) claims[cʃ] to be a more accurate rendering.
Moving away from the normative system, some languages and dialects have alveolar affricates[ts] instead of palatal, though some among them retain[tʃ] in certain positions: beforefront vowels (esp./i/), before/j/, or whengeminated. Alveolar as anadditional point of articulation occurs inMarathi andKonkani where dialect mixture and others factors upset the aforementioned complementation to produce minimal environments, in some West Pahari dialects through internal developments (*t̪ɾ,t̪ >/tʃ/), and inKashmiri. The addition of aretroflex affricate to this in someDardic languages maxes out the number of stop positions at seven (barring borrowed/q/), while a reduction to the inventory involves *ts >/s/, which has happened inAssamese,Chittagonian,Sinhala (though there have been other sources of a secondary/ts/), and Southern Mewari.
Further reductions in the number of stop articulations are in Assamese andRomani, which have lost the characteristic dental/retroflex contrast, and in Chittagonian, which may lose its labial and velar articulations throughspirantisation in many positions (>[f,x]).[47] /q x ɣ f/ are restricted to Perso-Arabic loanwords in most IA languages but they occur natively in Khowar.[48] According to Masica (1991) some dialects of Pashayi have a /θ/ which is unusual for IA languages. Domari which is spoken in the Middle East and had high contact with Middle Eastern languages has /q ħ ʕ ʔ/ and emphatic consonants from loanwords.
Marathi, Konkani, certain W. Pahari dialects (Bhadrawahi, Bhalesi, Mandeali, Padari, Simla, Satlej, maybe Kulu), Kashmiri, E. and N. dialects of Bengali (parts of Dhaka, Mymensingh, Rajshahi)
Hindustani, Punjabi, Dogri, Sindhi, Gujarati, Sinhala, Odia, Standard Bengali, dialects of Rajasthani (except Lamani, NW. Marwari, S. Mewari), Sanskrit,[49] Prakrit, Pali, Maithili, Magahi, Bhojpuri
Romani, Domari, Kholosi
Nepali, dialects of Rajasthani (Lamani and NW. Marwari), Northern Lahnda's Kagani, Kumauni, many West Pahari dialects (not Chamba Mandeali, Jaunsari, or Sirmauri)
Sanskrit was noted as having fivenasal-stop articulations corresponding to its oral stops, and among modern languages and dialects Dogri, Kacchi, Kalasha, Rudhari, Shina, Saurashtri, and Sindhi have been analysed as having this full complement of phonemic nasals/m//n//ɳ//ɲ//ŋ/, with the last two generally as the result of the loss of the stop from ahomorganic nasal + stop cluster ([ɲj] >[ɲ] and[ŋɡ] >[ŋ]), though there are other sources as well.[50]
In languages that lack phonemic nasals at some places of articulation, they can still occur allophonically from place assimilation in a nasal + stop culture, e.g. Hindi/nɡ/ >[ŋɡ].
Most Indo-Aryan languages have contrastiveaspiration (/ʈ/~/ʈʰ/), and some retain historicalbreathy voice on voiced consonants (/ɖ/~/ɖʱ/). Sometimes both phenomena are analysed as a single aspiration contrast. The places and manners of articulation which allow contrastive aspiration vary by language; e.g. Sindhi permits phonemic/mʱ/, but the phonemic status of this sound in Hindi is uncertain, and many "Dardic" languages lack aspirated retroflex sibilants despite having unaspirated equivalents.[51]
In languages that have lost breathy-voice, the contrast has often been replaced with tone.
Some of these are mentioned inMasica (1991:104–105).
Implosives: Languages in theSindhic subfamily, as well asSaraiki, westernMarwari dialects, and some dialects of Gujarati have developed implosive consonants from historical intervocalic geminates and word-initial stops. Sindhi has a full implosive series except for the dental implosive:/ɠʄᶑɓ/. It has been claimed thatWadiyari Koli has the dental implosive too. Other languages have less complete implosive series, e.g. Kacchi has just/ᶑɓ/.
Prenasalized stops: Sinhala and Maldivian (Dhivehi) have a series of prenasalised stops covering all places except for palatal:/ᵐbⁿdᶯɖᵑɡ/.
Palatalization: Kashmiri (natively) and some Romani dialects (from contact with Slavic languages) have contrastive palatalisation.
Voiceless lateral In Gawarbati, some Pashai dialects, partly Bashkarik and some Shina dialects have /ɬ/ from clusters of tr kr or sometimes pr; dr gr and br merged with /l/ in these languages.
Lateral affricates:Bhadarwahi has an unusual series of lateral retroflex affricates (/ʈ͡ꞎɖ͡ɭɖ͡ɭʱ/ derived from historical/Cɾ/ clusters.
Vowel typologies are varied across Indo-Aryan due to diachronic mergers and (in some cases) splits, as well as different accounts by linguists for even the widely-spoken languages. Vowel systems perMasica (1991:108–113) are listed below. Many languages also have phonemic nasal vowels.
W. Pahari (Dogri, Rudhari, Mandeali, Pangwali, Khashali, Churahi), Saraiki
/iɪe·aə·ɔoʊu/
W. Pahari (Shodochi, Surkhuli)
/iɪeɛ·a·ɔoʊu/
W. Pahari (Jaunsari, Shoracholi, Kullui)
8
/ieɛ·aə·ɔou/
Gujarati
/ieɛa·ɒɔou/
Assamese
/iɪe·aə·oʊu/
Halbi, Bhatri, W. Pahari (Garhwali, Chameali, Gaddi)
7
/ieæ·a·ɔou/
Bengali
6
/iea·ɔou/
Odia, Bishnupriya Manipuri
/ie·aə·ou/
Marathi, Nepali, Lambadi, Sadri/Sadani
5
/ie·a·ou/
Romani (European dialects)
Sylheti language is one of the fewtonal Indo-Aryan languages, others being Punjabi and a few Dardic languages. The vowels of Sylheti language listed below.[52]
The following are consonant systems of major and representative New Indo-Aryan languages, mostly followingMasica (1991:106–107), though here they are inIPA. Parentheses indicate those consonants found only in loanwords: square brackets indicate those with "very low functional load". The arrangement is roughly geographical.
In many Indo-Aryan languages, the literary register is often more archaic and utilises a different lexicon (Sanskrit or Perso-Arabic) than spoken vernacular. One example is Bengali's high literary form,Sādhū bhāṣā, as opposed to the more modernCalita bhāṣā (Cholito-bhasha).[55] This distinction approachesdiglossia.
In the context of South Asia, the choice between the appellations"language" and "dialect" is a difficult one, and any distinction made using these terms is obscured by their ambiguity. In one general colloquial sense, a language is a "developed" dialect: one that is standardised, has a written tradition and enjoyssocial prestige. As there are degrees of development, the boundary between a language and a dialect thus defined is not clear-cut, and there is a large middle ground where assignment is contestable.There is a second meaning of these terms, in which the distinction is drawn on the basis of linguistic similarity. Though seemingly a "proper" linguistics sense of the terms, it is still problematic: methods that have been proposed for quantifying difference (for example, based onmutual intelligibility) have not been seriously applied in practice; and any relationship established in this framework is relative.[56]
^Various counts depend on where the line is drawn between a "dialect" and a "language".[citation needed]Glottolog 4.1 lists 224 languages.
^Burde, Jayant (2004).Rituals, Mantras, and Science: An Integral Perspective. Motilal Banarsidass Publishers. p. 3.ISBN978-81-208-2053-1.The Aryans spoke an Indo-European language sometimes called the Vedic language from which have descended Sanskrit and other Indic languages ... Prakrit was a group of variants which developed alongside Sanskrit.
^Jain, Danesh; Cardona, George (26 July 2007).The Indo-Aryan Languages.Routledge. p. 163.ISBN978-1-135-79711-9.... a number of their morphophonological and lexical features betray the fact that they are not direct continuations of R̥gvedic Sanskrit, the main base of 'Classical' Sanskrit; rather they descend from dialects which, despite many similarities, were different from R̥gvedic and in some regards even more archaic.
^Chamber's Encyclopaedia, Volume 7. International Learnings Systems. 1968.Most Aryan languages of India and Pakistan belong to the Indo-Aryan family, and are descended from Sanskrit through the intermediate stage of Prakrit. The Indo-Aryan languages are by far the most important numerically and the territory occupied by them extends over the whole of northern and central India and reaches as far south as Goa.
^Standard Hindi first language: 260.3 million (2001), as second language: 120 million (1999). Urdu L1: 68.9 million (2001–2014), L2: 94 million (1999):Ethnologue 19.
^Bengali or Bangla-Bhasa, L1: 242.3 million (2011), L2: 19.2 million (2011),Ethnologue
^Eberhard, David M.; Simons, Gary F.; Fennig, Charles D., eds. (2020).Ethnologue: Languages of the World (23rd ed.). Dallas, Texas: SIL International.
^Hammarström, Harald; Forkel, Robert; Haspelmath, Martin, eds. (2017)."Indo-Aryan".Glottolog 3.0. Jena, Germany: Max Planck Institute for the Science of Human History.
^Kogan, Anton I. (2005).Dardskie yazyki. Geneticheskaya kharakteristika [Dardic language. Genetic characteristic] (in Russian). Moskva: Vostochnaya literatura.
^Cardona, George; Jain, Dhanesh, eds. (2003). "The historical context and development of Indo-Aryan".The Indo-Aryan Languages. Routledge language family series. London:Routledge. pp. 46–66.ISBN0-7007-1130-9.
^Claus, Peter J.; Diamond, Sarah; Mills, Margaret Ann (2003). "Afghanistan, Bangladesh, India".South Asian folklore: an encyclopedia.Routledge. p. 203.
^Hammarström, Harald; Forkel, Robert; Haspelmath, Martin, eds. (2017)."Tharuic".Glottolog 3.0. Jena, Germany: Max Planck Institute for the Science of Human History.
^Hammarström, Harald; Forkel, Robert; Haspelmath, Martin, eds. (2017)."Kuswaric".Glottolog 3.0. Jena, Germany: Max Planck Institute for the Science of Human History.
^Hammarström, Harald; Forkel, Robert; Haspelmath, Martin, eds. (2017)."Chinali–Lahul Lohar".Glottolog 3.0. Jena, Germany: Max Planck Institute for the Science of Human History.
^Paul Thieme, The 'Aryan' Gods of the Mitanni Treaties. JAOS 80, 1960, 301–17
^Parpola, Asko (2015).The Roots of Hinduism: The Early Aryans and The Indus Civilization.Oxford University Press.
Morgenstierne, Georg. "Early Iranic Influence upon Indo-Aryan." Acta Iranica, I. série, Commemoration Cyrus. Vol. I. Hommage universel (1974): 271–279.
John Beames,A comparative grammar of the modern Aryan languages of India: to wit, Hindi, Panjabi, Sindhi, Gujarati, Marathi, Oriya, and Bangali. Londinii: Trübner, 1872–1879. 3 vols.
Madhav Deshpande (1979).Sociolinguistic attitudes in India: An historical reconstruction. Ann Arbor: Karoma Publishers.ISBN0-89720-007-1,ISBN0-89720-008-X (pbk).
Kobayashi, Masato.; &George Cardona (2004).Historical phonology of old Indo-Aryan consonants. Tokyo: Research Institute for Languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies.ISBN4-87297-894-3.
Sen, Sukumar. (1995).Syntactic studies of Indo-Aryan languages. Tokyo: Institute for the Study of Languages and Foreign Cultures of Asia and Africa, Tokyo University of Foreign Studies.
Vacek, Jaroslav. (1976).The sibilants in Old Indo-Aryan: A contribution to the history of a linguistic area. Prague: Charles University.
