Languages and Scripts
LANGUAGES AND SCRIPTS
LANGUAGES AND SCRIPTS The Indian subcontinent has been a virtual laboratory of humanity, from its first settlement by modern humans while on their way to Australia and East Asia, some 50,000 years ago. The remnants of this founding population can still be found in genetics, linguistically in some of the "tribal" languages (Andamanese, Kusunda), and in various substrates. The initial settlement was followed by the immigration of speakers of diverse languages belonging to a number of major language families.
India's language families are, from north to south: Tibeto-Burmese, Indo-European (Iranian and Indo-Aryan), Khamti (Tai), Austro-Asiatic (Munda, Khasi), and Dravidian. There also are several isolates. Burushaski is regarded by some as the eastern extension of the Macro-Caucasian (Basque–N. Caucasian–Burushaski) family; Kusunda and Andamanese are perhaps linked to Indo-Pacific (Papuan, Tasmanian). In the past, many more languages must have been spoken, as the little studied substrates indicate. Substrates are found in Nahali, Tharu, Vedda, and even in Hindi, where some 30 percent of the agricultural terms are from an unknown source. The same seems to be true for many other North Indian languages; southern Indian substrates have hardly been touched as yet. In sum, South Asia was as linguistically diverse then as it is today.
Most South Asian languages now belong to the Indo-Aryan subgroup of the Indo-European language family that stretched, before the colonial period, from Iceland to Bengal and Sri Lanka. Speakers of Old Indo-Aryan (OIA) must have entered South Asia in the mid-second millennium b.c. OIA, beginning with the Vedic form of Sanskrit, is a close relative of Old Iranian, which has been preserved in some Old Persian inscriptions (519 b.c.) and in the Avesta, the sacred language of the Zoroastrians (Parsi). Dating the earliest OIA is difficult; however, the Vedic dialect that was brought into the Mitanni realm of northern Iraq/Syria (c. 1400 b.c.) is slightly more archaic than the oldest Vedic text in India, the Rig Veda. The early forms of Sanskrit show clear substrate influences, in part from the Bactria-Margiana Archaeological Complex (2400–1600 b.c.) and from the Hindu Kush–Pamir areas, all of which points, together with the Mitanni, to evidence of an influx of OIA speakers into South Asia from the northwest.
Other forms of OIA include middle and late Vedic; Classical Sanskrit, still a fully inflectional language like Greek and Latin, emerged from a conservative form of OIA. It was codified by the ingenious grammarian Pāṇini, who lived near Attock on the Indus (c. 400 b.c.). Vedic OIA was spoken from this area up to the borders of Bengal and also in western Madhya Pradesh and northern Maharashtra. In its classical form (Sanskrit) it expanded, as the language of the learned, all over South Asia and beyond, to Bali, Vietnam, and, with Buddhist missions, through Afghanistan into Central Asia, China, and Japan.
By the time of the Buddha (c. 400 b.c.) OIA had been supplanted as popular language by Middle Indo-Aryan (MIA); some MIA forms are occasionally visible in Vedic texts. The Buddha taught in MIA, though texts in his own eastern dialect have not been preserved; instead, they have come down in the western Pali variety of MIA. Fragments of Buddhist texts also survive in some more recent forms of MIA, as in the oldest Buddhist manuscripts from the first century a.d. The first datable testimony of MIA are the inscriptions of the emperor Ashoka (c. 250 b.c.) They are found, in various dialect forms, all over South Asia. They are, however, in Greek and Aramaic in western Afghanistan, and they are absent in the deep (Dravidian) south.
MIA differs from OIA by a number of phonetic and grammatical innovations, such as the assimilation of consonant groups and the restructuring of the verb system. MIA has heavily influenced the epic Sanskrit of the Mahābhārata and the Rāmāyaṇa. Around the beginning of the common era, several later Prakrit forms of MIA emerged that were used in classical Sanskrit dramas: Śaurasenī in the midlands, Māhārāstrī in southwest India, the little used "popular" Māgadhī in the east; further, the little known Paiśācī, the Ardha-Māgadhī preferred by the Jains, and the Gāndhārī of some Buddhist texts. For a millennium, Prakrit languages were used in inscriptions and poetry as well as in Jain texts, until they gradually gave way to early forms of New Indo-Aryan (NIA) during the latter part of the first millennium a.d. Further loss of grammatical endings is a characteristic of NIA; therefore, both the noun and verb systems were completely restructured, somewhat along the lines of the change from Latin to the Romance languages.
The earliest form of NIA is the Apabhraṃśa, preserved in quotes recorded by medieval writers and in the Buddhist stanzas in an eastern dialect. Most of the religious poets of the High Middle Ages used early NIA. Apabhraṃśa and its later form Avahat.ṭha developed into the modern NIA languages during the second millennium a.d. These languages can be divided into central languages (various forms of Hindi) and outlying ones, or into some eight separate subgroups: the conservative northwestern Dardic (such as Kalash, Khowar, Shina, Kashmiri); the partially conservative Pāhārī (in Himachal Pradesh, Uttarkhand, and Nepālī, which is spoken up to Sikkim, Bhutan, and Assam); the languages of the Punjab (Lahnda, Panjabi) and Sind, of western India (Gujarati, Marathi, with its Konkani dialect in Goa); the southernmost languages Sinhala (Sri Lanka) and Dhivehi (Maldives); the eastern group with Maithili and Magahi (Bihar), Bengali, Oriya, and Assamese. The large central part of North India is covered by various forms of Hindi. Next to the standard language, there are Bhrāj, the more divergent Bhojpuri, and Sadani, or Nimari in the South. Urdu is a variant of central NIA and virtually the same language as Hindi, though with much more Persian and Arabic vocabulary; it is written in a version of the Arabic script. Consequently, Pakistan and Kashmir use Urdu, and most states of North India, from Himachal Pradesh up to Andhra and West Bengal, use Hindii.
Compared to the dominance of NIA languages and their speakers, the Dravidian languages are by and large limited to South India, where Tamil, Malayalam, Telugu, Tulu, and Kannada are the dominant languages today. Its earliest forms are attested, from about 200 b.c. onward, in the Tamil inscriptions of Tamil Nadu and Kerala, and in the classical Tamil texts of the Sangam poems. The early inscriptions indicate influences from a lost proto-Kannada. The other Dravidian languages have been attested only hundreds of years later, and many "tribal" ones only over the past two hundred years. The time and exact location of the influx of Dravidian into the subcontinent is unclear, but there are clear remnants in topographical terms in Sind, Gujarat, and Maharashtra. These and early cultural influences on Dravidian all point westward: agricultural words came from Sumerian or Elamite, and some even from farther west. However, the language(s) of the Indus Civilization (2600–1900 b.c.) are uncertain, as its inscriptions cannot be read (if indeed they are a script). The supposed linguistic remnant of that time, modern Brahui in Baluchistan, however, is a late medieval immigrant from central India (as is Kurukh/Oraon and Malto in Jharkhand). Central India has also retained various Goṇḍī dialects, and South India, especially the Nilgiris (with Toda), has other "tribal" Dravidian languages. A conservative dialect of Tamil has been spoken in northern and northeastern Sri Lanka for many centuries and has considerably influenced the Indo-Aryan Sinhala.
Proto-Dravidian is an agglutinative language like its supposed relative, Uralic. It shares many areal features with Central, North, and Northeast Asia and, surprisingly, highland Ethiopia. Some scholars unite Dravidian and Indo-European, along with South Caucasian (Georgian), Afro-Asiatic (Semitic, Old Egyptian, Berber, etc.), Uralic, and Altaic (including Korean and Japanese) in a Nostratic superfamily or, according to a recent proposal, in Eurasiatic. The historical development of the Dravidian languages has been investigated only on a limited basis thus far, in spite of its importance both for the history of the subcontinent as well as for the little-studied southern influence on the literary and religious history of the North. There also has been an enormous influence of Dravidian on the northern languages, starting with certain parts of the oldest OIA text, the Rig Veda. While its earliest stages seem to lack Dravidian words, the text otherwise has a number of early Dravidian loans. Their amount increases throughout Vedic literature, and their impact continues until today, though most speakers of Sanskrit or NIA do not recognize that such words as daṇḍa (stick) are loans from Dravidian. Strong Dravidian influence is also seen in phonology, word formation, and syntax of OIA, MIA and NIA: we can therefore speak of a South Asian "linguistic area" (sprachbund). This linguistic area also includes the Munda languages.
Northern Munda is now spoken in eastern and also in western India (Korku, on the Tapti River). In the east, Santali (West Bengal, Jharkhand) and Mundari are the languages with the most speakers; however, Northern Munda (Kherwari) also includes Asuri, Ho, and so on; Central Munda includes Kharia, Juang, and others; and Southern Munda, spoken on the borders of Orissa and Andhra, includes Sora, Pareng, Gutob, and others. Munda itself is the western branch of Austro-Asiatic, which includes, inside India, also Khasi and War (Meghalaya) as well as Nicobar, and outside India, Mon-Khmer, the related Tai languages, and Vietnamese. Khamti (northern Tai) was the language of the medieval Ahom kingdom and still survives in northeastern Assam. Austro-Asiatic must once have been spoken in a much larger area of North India, as many prefixing loans (often closer to War-Khasi) survive in the Rig Veda. Proto-Munda, which has been reconstructed only over the past few decades, was a largely mono/bisyllabic language working with prefixes and infixes. Under the sprachbund influence of Indo-Aryan, it has shifted to a (largely) suffixing language, especially Northern Munda, which also has an enormous number of NIA loans. The largely uninvestigated Munda and War/Khasi languages (and their oral literatures) contain much that is important for the linguistic, religious, and cultural prehistory of the subcontinent. In the Jharkhand and Meghalaya states, at least Santali/Mundari and Khasi may be maintained, while many of the small tribal languages face quick extinction due to the pressure of neighboring NIA and Dravidian languages.
The northern rim of South Asia is occupied by Tibeto-Burmese languages that are linked to Chinese (Sino-Tibetan). Their early history in the subcontinent is unknown; however, even early Vedic texts speak of a mountain people, the Kirāta. The earliest documentation is found in place names and in the Licchavi inscriptions (Kathmandu Valley, c. a.d. 200–750), which also refer to the Kirāta. After a.d. 983, there are land sale inscriptions on palm leaf in early Newari, the dominant language of the Kathmandu Valley until the Gorkha conquest in 1768–1769, when NIA Nepālī became the offical language. Most other Tibeto-Burmese languages of the Himalayan belt have not left records until very recently. From west to east, they include among others: Magar, Gurung, Tamang, Rai, Limbu, Lepcha, Bodo, Naga, Meitei (in Manipur). In the northernmost Himalayan uplands, various forms of Tibetan are spoken, such as the archaic western dialect of Ladakh, the eastern one of the Sherpa, or the southern one of Bhutan (now the official language, Dzonkha). Early influence of Tibeto-Burmese on OIA can be discerned in the Vedas, such as the topographical names Kosi and Kosala, or the word for cooked rice (Hindi, cawal; Nepālī, camal).
The rest of the South Asian languages, whether remnants or substrates, have been recorded only over the past two centuries. Taken together, they open a wide vista of early relationships: westward to the Caucasus (seen in Burushaski, Dravidian), to Central Asia and beyond (Vedic Sanskrit, Indo-Iranian), Southeast Asia (origin of Munda, Khasi, Khamti), and Greater Tibet (Tibeto-Burmese). Even wider relationships emerge in the remnant languages and substrates (Kusunda with Andamanese/Papuan, Nahali with Ainu, which is now also evident in genetic data, i.e., sharing the early Y chromosome haplogroup IV videlicet D). The predominance of retroflex sounds found from the Hindu Kush east and southward perhaps was a feature of even the earliest substrates (as in Andamanese).
In spite of the great diversity of the prehistoric, ancient, and modern South Asian languages, it has been the Vedic form of Sanskrit (still used in ritual) and especially Classical Sanskrit that have exercised a dominant influence on the languages of the subcontinent. Sanskrit has functioned as the language of administration, scholarly discourse, and religion, not unlike Latin in Europe. In spite of the dominant role of Persian in most of northern India after a.d. 1200–1500, Sanskrit remained the language even of some diplomatic and administrative contacts well into the British period. According to census figures (1991), some 45,000 people still claimed it as their (near-)native tongue. Like Latin and Greek in the West, Sanskrit continues to supply technical terms for administration (rāṣṭrapati, "president"; antārāṣṭrīya, "international") or technology (ākāśvāṇī, "radio"; dūrdarśana, "television"; jala vidyut āyoga, "hydel (or hydro-electric) project"). Sanskrit influence on the "officialese" of most modern languages is so strong that one jokes that one cannot listen to "the news in Hindi" but rather has to look for "the Hindi in the news." There is a stealthy Sanskritization of most South Asian languages.
However, since the British colonial period, and now with globalization, English has increasingly become the lingua franca, largely supplanting the official language, Hindi. South Asian English has developed into a distinct dialect with strong substrate influences, such as pronunciation of dentals (t, d, th) as retroflexes, a peculiar pitch intonation with partially misplaced accents, lack or hypercorrect insertion of the definite article, colloquial changes in the verbal system ("I am knowing this"), and an abundance of substrate words (dacoit, "bandit"; godown, "storage") and frequent code switching with the local language.
The earliest scriptlike symbols of the subcontinent belong to the Harappan or Indus Valley Civilization (2600–1900 b.c.). They remain undeciphered, though they have been claimed to represent an early form of Dravidian, and though it is entirely uncertain whether Dravidian was indeed spoken there and to what extent. The substrates in Vedic rather point to a number of Indus languages. In addition, there even is no consensus about the exact number of the Indus signs (400–600), as to their mutual combinations, and whether the signs represent an alphabet, a syllabo-logographic script, or no script at all. In spite of some regional differences, however, the signs are fairly well standardized throughout the large Harappan area (Pakistan, Punjab, Haryana, and Gujarat). Whether the symbols were used in trade or to indicate political dominance, they quickly disappeared when the Indus Valley Civilization disintegrated. Attempts to link these symbols with later scripts are doomed. Barring a decipherment, the Indus signs provide enough simple, often geometric forms that can be connected with any early script, or even with some potter's and mason's marks. Further, there is a millennia-long gap between the last Indus inscriptions (c. 1900 b.c.) and the first ones in later Indian scripts (under Emperor Ashoka, c. 250 b.c.).
The first writing in South Asia appears only after the impact of Persian domination of much of Pakistan (530 or 519–327 b.c.). Indeed, Karoṣṭhī, the earliest script of the northwestern subcontinent, is an Aramaic-based, rather cursive, Semitic-style alphabet. Aramaic was the language of administration in the Persian empire and even in some of Ashoka's inscriptions (in Kandahar and Taxila). Unlike all other Indian scripts, Kharoshthi is written from right to left. It is a true alphabet: all vowels that begin a syllable and all consonants are represented by individual signs, but vowel length is not marked. However, due to the abundance of a/ā sounds in MIA, postconsonantal a/ā are not written but are inherent in the consonant sign. All other postconsonantal vowels are marked by small diacritical signs above, below, or crossing the consonant sign in question. Certain consonant clusters are expressed by ligatures or special individual signs. Unlike Semitic alphabets such as Aramaic, the script therefore is well attuned to the Indian phonetic system. Kharoshthi is older than its sister script, Brahmi, and continued to be used until the third century a.d.in the northwest and in Central Asia (Xinjiang).
The other script used in Ashoka's inscriptions in the rest of India was Brahmi. It seems to have been derived from Semitic scripts (Aramaic, Kharoshthi), but it was completely reconfigured, perhaps under contemporary Greek (Hellenistic) influence, which would be plausible if Brahmi was indeed created under Ashoka. Like Kharoshthi, Brahmi perfectly fits the various contemporary MIA languages; however, it clearly was not designed for Sanskrit. Several Sanskrit phonemes or their allophones (such as ) that are missing in MIA are not represented, and Sanskrit inscriptions begin only in the first century b.c. (Ayodhya, Mathura). Brahmi also lacks a method to mark the final vowel-less consonants of Sanskrit words, a feature that does not occur in MIA. As in Kharoshthi, short postconsonantal -a is not written but inherent in the consonant sign. However, long ā and all other postconsonantal vowels are marked by small diacritical signs above, below, and to the right of the consonant. Only superficially, this system may look like a mixture of an alphabet with a syllabary. With little variation, the system has been followed until today (except in the early Tamil Brahmi script). Consonant groups are mostly represented by writing a single consonant (in MIA), and later on by a ligature with one subscribed consonant below the other (for Sanskrit). The perfect match between MIA/Sanskrit phonemes and the script has often been explained as the influence of the well developed Vedic phonetic sciences and of Pānini's grammar. Indeed, Indian alphabets follow, unlike the Semitic ones, a strictly phonetic arrangement, beginning with vowels, then with the consonants arranged from their place of realization at the back of the mouth (velars) to the front (labials), followed by resonants and sibilants.
Recently, the finds of some small fragments of inscribed pot shards in Sri Lanka (and then in Tamil Nadu) have cast some doubt about the age of the Brahmi script. Various early dates have been claimed for the finds, all of which are pre-Ashoka. However, small pottery shard fragments can easily be transported through rat holes into lower archaeological levels. More finds are to be awaited. The Brahmi inscriptions must also be viewed in the context of early Tamil Brahmi used in southernmost India (Tamil Nadu and Kerala). Since the second century b.c., Brahmi has been used to write early Tamil, with four signs added for sounds restricted to Tamil. Other than in Ashokan Brahmi, a short postconsonantal -a and all other vowels were indicated by a diacritic, thus k+a, k+ā, k+i, and so forth. Thus, unlike in all other Indian scripts, double consonants could be written simply by doubling the sign. It is unclear whether this system was based on Ashokan Brahmi that was used in nearby Karnataka, Andhra, and Sri Lanka, or whether it went back to a lost early South Indian form. In sum, Brahmi was used, with little regional variation, all over South Asia, with the exception of the northwest (Kharoshthi). All later alphabets of South Asia (and most of Southeast Asia and Tibet) are derived from Ashokan Brahmi. Even the ordering of the native Japanese Kana syllabary is based on the Indian alphabetic order. During the first centuries a.d., the Brahmi script gradually developed nail-like extensions at the top of the letters (head markers) and, due to writing with ink on palm leaves, a more cursive form as well as some regional variations. A few letters were added that were necessary to write Sanskrit (ṇ, ḥ, intervocalic l̄, allophones of ḥ, h̄, h̆). A diacritic indicating vowel-less consonants was added, as well as increasingly more conjuncts (ligatures) for consonant groups (a system used until today). Also, an (invisible) square frame was gradually developed for all letters, which resulted in the squarish northern Gupta script (due to use of ink and pen, c. a.d. 300) and the more rounded southern variety (due to use of stylus for incising letters on palm leaf).
The intricacies of paleographical development cannot be traced here. However, the early split between northern and southern alphabets increased. The northern Gupta script developed, during the sixth century, into the angular Siddhamātrikā script that was widely used—even, due to the spread of Buddhism, in China and Japan (where it still survives as Siddham script). A subvariety emerged in the northwest: the early Śāradā script, used in Kashmir, Himachal Pradesh, Gandhara, Swat, and the Punjab, where modern Gurmukhi is a distant descendant. Śāradā slowly developed over the Middle Ages and was used in Kashmir by Brahmans well into the twentieth century. Another variety developed in the east: Eastern Nagari (or Proto-Bengali, Gauḍī). It took local forms in Bengal, Mithila, and in the Kathmandu Valley (Newari script with many attested subsequent forms: Bhujimol/Kuṭilā/Rañjanā, etc.) Modern Newari script is still used for ceremonial purposes. Early Bengali script developed, around 1400, into the Oriya script that favored, as in South India, round shapes of letters.
The mainstream Siddhamatrika developed, around a.d. 1000, into early Devanāgarī, with its typical horizontal top line (not maintained in all Siddhamatrika-derived scripts). It has been used in the central area, but also in Gujarat, where it was used by Jains and Brahmans well into the nineteenth century, when it developed its modern Gujarati forms separately. Devanāgarī has also been used in Maharashtra. In the middle ages, a somewhat variant form (Nandināgarī) was used in the southern Vijayanagara and Tanjore kingdoms. Due to the selection of Hindi as India's official language, the use of Nagari has spread to the Himalayas, Haryana, Rajasthan, Madhya Pradesh, Chattisgarh, Jharkhand, and Bihar. Sanskrit publications everywhere now use Devanāgarī, though Sanskrit, which has always been written in local scripts, is still written in local forms for local use.
In South India, scripts deviated at the same time as in the North. The Pallava script of Tamil Nadu (c. a.d. 500) also had a profound influence on the development of all Southeast Asian scripts. In Karnataka and Andhra, the rounded southern characters became almost fully closed, resulting in the modern Telugu-Kannada scripts. In the deep South, the Grantha script (for Sanskrit) and Tamil were developed from earlier southern scripts. They retain, to some extent, the less rounded forms of the late South Brahmi and Pallava scripts. Tamil has the shortest Indian alphabet, due to an ingenious system of writing only phonemes (neglecting predictable allophones, thus k [k, x]) and because of the lack of ligatures, as vowel-less consonants are marked with a superscripted dot/circle (puḷḷi). Thus, Tamil script has been easiest to learn, a feature of Tamil writing systems from the beginning. In the middle of the second millennium, Malayalam script was developed out of the Grantha script. The Sinhala script of Sri Lanka is another development of southern Brahmi script. It has been influenced by other medieval South Indian scripts while achieving its typical, rounded modern form.
The Siddhamatrika-derived Tibetan script, close to early Sharada, is used in Tibet and in the northernmost areas of the subcontinent, as well as for the national language of Bhutan (Dzongkha). The Limbu, Lepcha, and Meitei scripts are based on a version of the Tibetan script and were used for religious writings. One recently developed form of the Meitei script has been revived as the quasi-official script of Manipur.
All Indian scripts are unique and perfect adaptations to the Indian sound systems (phonemes, including some allophones). Most of them, however, remain unwieldy (even in computer use) due to their heavy reliance on individual ligatures for consonant clusters, by positioning Nagari i- before consonants, or due to the split up of the signs for medial -e-, -o- (Bengali, Orissa, southern alphabets). Similarly, the Arabic script used for Urdu and some regional languages (Khowar, Shina, Kashmiri) remains inadequate in expressing the complex vowel systems of NIA. One can only read the Urdu script well if one knows the words intended. However, in Bangladesh the Bengali alphabet (and many Sanskrit loanwords) have been retained. Persian language and script were widely used during the Middle Ages and in the early British colonial period, until Persian was replaced by English in 1835.
The English alphabet has been used since for a variety of goals, such as street signs or film advertisements. Some tribal languages (and Nepālī as used by the British army) have also been written in the Roman alphabet, though Devanāgarī has been introduced in many areas more recently. After independence, some had proposed to make a variety of the Roman alphabet the national script, but this idea did not take hold. The recent adaptations for computer use have given a further boost to South Asian alphabets, though the quick spread of computers has also substantially increased the use of the English language, which is still spoken by only a tiny minority of South Asians.
Anderson, G. D. S. "Recent Advances in Proto-Munda and Proto-Austroasiatic Reconstruction." Paper at the Third Round Table on the Ethnogenesis of South and Central Asia. Available at <http://www.fas.harvard.edu/%7Esanskrit/images/html_images/Andersonaamorph-rtf.pdf>
Benedict, P. K. Sino-Tibetan: A Conspectus. Cambridge, U.K.: Cambridge University Press, 1972.
Berger, H. Die Burushaski-Sprache. Wiesbaden: Harrassowitz, 1999.
Cardona, George. Panini, His Work, and Its Traditions. Delhi: Motilal Banarsidass 1988.
Emeneau, M. B. "India as a Linguistic Area." Language 32 (1956): 3–16.
——. Dravidian Borrowings from Indo-Aryan. Berkeley: University of California Press, 1962.
——. Language and Linguistic Area: Essays, Selected and Introduced by Anwar S. Dil. Stanford, Calif.: Stanford University Press, 1980.
——. Toda Grammar and Texts. Philadelphia: American Philosophical Society, 1984.
Farmer, S., R. Sprout, and M. Witzel. "The Collapse of the Indus-Script Thesis: The Myth of a Literate Harappan Civilization." Electronic Journal of Vedic Studies 11, no. 2 (2004): 19–57. Available at <http://users.primushost.com/~india/ejvs/issues.html>
Hansson, G. The Rai of Eastern Nepal, Ethnic and Linguistic Grouping: Findings of the Linguistic Survey of Nepal. Kathmandu: Linguistic Survey of Nepal and Centre for Nepal and Asian Studies, Tribhuvan University, 1991.
Hart, G. L. Poets of the Tamil Anthologies: Ancient Poems of Love and War. Princeton, N.J.: Princeton University Press, 1979.
Hinüber, O. von. Das ältere Mittelindisch im Überblick. Vienna: Verlag der Æsterreichischen Akademie der Wissenschaften, 1986. 2nd rev. ed., 2001.
Hultzsch, E. Inscriptions of Asoka. Corpus inscriptionum Indicarum, vol. 1. Reprint, Delhi: Indological Book House, 1969.
Kölver, B., and U. Kölver. "On Newārī noun inflection." Zentralasiatische Studien des Seminars für Sprach- und Kulturwissenschaft Zentralasiens der Universität Bonn 9 (1975): 87–117.
——. "Classical Newārī verbal morphology." Zentralasiatische Studien des Seminars für Sprach- und Kulturwissenschaft Zentralasiens der Universität Bonn 12 (1978): 273–316.
Kuiper, F. B. J. "The Genesis of a Linguistic Area." Indo-Iranian Journal 10 (1967): 81–102.
Lubotsky, A. "The Indo-Iranian Substratum." In Early Contacts between Uralic and Indo-European: Linguistic and Archaeological Considerations, edited by C. Carpelan et al. Helsinki: Suomalais-Ugrilainen Seura, 2001.
Mahadevan, I. Early Tamil Epigraphy: From the Earliest Times to the Sixth CenturyA.D. Chennai: Cre-A; Cambridge, Mass.: Harvard University Press, 2003.
Masica, C. P. "Aryan and Non-Aryan Elements in North Indian Agriculture." In Aryan and Non-Aryan in India, edited by M. Deshpande and P. E. Hook. Ann Arbor: Center for South and Southeast Asian Studies, University of Michigan, 1979.
——. The Indo-Aryan Languages. Cambridge, U.K., and New York: Cambridge University Press, 1991.
Rajan, V.S. A Reference Grammar of Classical Tamil Poetry: 150 b.c.–Pre-Fifth/Sixth Century a.d. Philadelphia: American Philosophical Society, 1992.
Salomon, R. Indian Epigraphy: A Guide to the Study of Inscriptions in Sanskrit, Prakrit, and the Other Indo-Aryan Languages. New York: Oxford University Press, 1998.
Sivaramamurti, C. "Indian Epigraphy and South Indian Scripts." Bulletin of the Madras Government Museum, New Series, General Section, 3, no. 4. Chennai: Government of Madras, 1948; reprint, 1966.
Steever, S. B., ed. The Dravidian Languages. London: Routledge, 1998.
Szemerényi, O. Introduction to Indo-European Linguistics. Oxford: Clarendon Press; New York: Oxford University Press, 1996.
Tagare, G. V. Historical Grammar of Apabhramsa. Poona: Deccan College Post-graduate and Research Institute, 1948.
Tulpule, S. G., and A. Feldhaus. A Dictionary of Old Marathi. Oxford: Oxford University Press, 2000.
Westhouse, P., T. Usher, M. Ruhlen, and W. S.-Y. Wang. "Kusunda: An Indo-Pacific Language in Nepal." Proceedings of the National Academy of Sciences (2004) 101: 5692–5695.
Witzel, M. "Tracing the Vedic Dialects." In Dialectes dans les littératures indo-aryennes, edited by Colette Caillat. Paris: Collège de France, 1989.
——. "Early Sources for South Asian Substrate Languages." Mother Tongue, October 1999.
——. "Linguistic Evidence for Cultural Exchange in Prehistoric Western Central Asia." Sino-Platonic Papers 129 (2003).
Zide, N. H. "Munda and Non-Munda Austroasiatic Languages." Current Trends in Linguistics 5 (1969): 411–430.