Indo-Aryan languages

• 1. Introduction

  ›

  The Indo-Aryan languages are spoken by roughly 800 million people as of the early 21st century, making them one of the great language families on earth. They stretch across Eastern Pakistan, Northern India, southern Nepal, Bangladesh, Sri Lanka, and the Maldives. Yet their reach does not stop at the subcontinent. Romani speakers carry an Indo-Aryan tongue through the streets of Southeastern Europe, and communities spread from Northwestern Europe to Polynesia keep these languages alive far from where they were born.

  There are in the vicinity of 200 Indo-Aryan languages in total. Some estimates of total speakers climb as high as 1.5 billion. Hindustani, spoken by around 330 million people as a first language, sits at the top. Bengali follows with 242 million, Punjabi with about 150 million, Marathi with 112 million, and Gujarati with 60 million.

  How did a single ancestral tongue become so vast and so varied? The answer runs back more than three thousand years, to a reconstructed ancestor language that scholars call Proto-Indo-Aryan, and forward through centuries of conquest, migration, and literary flourishing that shaped everything from the Vedas to the national anthems of two modern nations.

• 2. Deep Roots and Ancient Evidence

  ›

  Proto-Indo-Aryan is a reconstructed language, meaning no texts survive from it directly. Linguists have worked backwards from its descendants to piece together what it must have looked like before 1500 BCE. Its closest attested relative is Vedic Sanskrit, the language of the oldest preserved Hindu sacred texts.

  The most striking early evidence appears far from South Asia, in the ruins of the Mitanni civilisation of Upper Mesopotamia. Around 1400 BCE, the Mitanni left traces of an Indo-Aryan language embedded in their records. While the Mitanni wrote mainly in Hurrian and Akkadian, their proper names and specialist vocabulary carry clear Indo-Aryan features. A treaty between the Hittites and the Mitanni invokes the deities Mitra, Varuna, Indra, and the Ashvins, names drawn straight from Vedic tradition.

  A horse-training text from the same civilisation, attributed to a figure named Kikkuli, records technical terms that match Sanskrit numerals almost exactly: aika for eka meaning "one", tera for tri meaning "three", panza for panca meaning "five", and satta for sapta meaning "seven". The numeral aika is particularly telling because it distinguishes the Mitanni connection from broader Indo-Iranian influence; early Iranian languages use aiva for "one", not aika. Scholars have also decoded Mitanni royal names in Sanskritic terms, reading Tushratta as carrying the sense of "whose chariot is vehement", and Citrarata as "whose chariot is shining". These traces are significant enough to push the known boundary of Indo-Aryan history back toward the Bronze Age Near East.

• 3. Sanskrit, Prakrit, and the Great Divergence

  ›

  Vedic Sanskrit grew into Classical Sanskrit over centuries of refinement. The word Sanskrit itself literally means "put together, perfected, elaborated", and it became the prestige language of culture, science, religion, and the court. By the time of the later Vedic texts, Sanskrit had shifted far enough that it was largely mutually unintelligible with the older Vedic form.

  Outside the learned sphere, ordinary people kept speaking vernacular dialects called Prakrits. The oldest surviving Prakrits are Pali, the canonical language of Buddhism, and Ardhamagadhi Prakrit, used in Jain scripture. Ashoka, the great Maurya emperor, inscribed his edicts in Ashokan Prakrit across the subcontinent, seeding a tradition of monumental written language.

  By the medieval period, the Prakrits had fragmented and evolved further. The transitional dialects of the roughly 6th to 13th centuries go by the umbrella term Apabhramsha. One text from this period stands out: the Shravakacara of Devasena, dated to the 930s, is now considered to be the first book written in Hindi.

  The Muslim conquests of the 13th through 16th centuries shifted the landscape again. Under the Mughal Empire, Persian became the language of prestige in the Islamic courts. This infusion of Persian, layered onto Apabhramsha foundations, drove the emergence of both Urdu and the various forms that eventually became modern Hindi. From that same Apabhramsha pool came Bengali, Bhojpuri, Hindustani, Assamese, Sindhi, Gujarati, Odia, Marathi, and Punjabi.

• 4. Hindustani and the Partition

  ›

  Braj Bhasha held the position of prestige dialect across the Central Zone Hindi-speaking areas for a long time. In the 13th century, it was displaced by Dehlavi-based Hindustani. Persian influence shaped that new prestige form steadily, and later waves of Sanskrit influence produced the registers we now call Modern Standard Hindi and Modern Standard Urdu.

  For centuries these two registers coexisted within the same political space. The division of the British Indian Empire in 1947 gave each a separate national role: Modern Standard Hindi became the official language in India, while Modern Standard Urdu became official in Pakistan. Their scripts differ, Devanagari for Hindi and a Perso-Arabic script for Urdu, but the underlying grammar remains identical. Linguists treat the difference as more sociolinguistic than purely linguistic.

  Today, Hindustani is the third most-spoken language in the world in terms of total speakers, counting both registers together. It functions as a second or third language for many more people across South Asia, a practical lingua franca across a subcontinent of extraordinary linguistic variety.

• 5. Languages That Left the Subcontinent

  ›

  Bengali is the seventh most-spoken language in the world and carries a rich literary tradition. The national anthems of both India and Bangladesh are written in Bengali, a fact that places the Eastern Indo-Aryan zone at the symbolic heart of two nations.

  The Eastern Indo-Aryan languages as a group share a notable feature: they display morphosyntactic similarities to Munda languages, which are largely absent from western Indo-Aryan. Scholars suggest that proto-Munda languages once dominated the eastern Indo-Gangetic Plain and were absorbed by Indo-Aryan as it spread eastward.

  Not all Indo-Aryan migrations stayed within the subcontinent. Domari is spoken by Dom people scattered across the Middle East, reaching as far north as Azerbaijan and as far south as central Sudan. Parya survives in Tajikistan and Uzbekistan, retaining features close to Punjabi and Western Hindi while absorbing influence from Tajik Persian. The Sindhic languages Kholosi, Jadgali, Luwati, Maimani, and Al Sayigh established themselves in the Persian Gulf region, probably through sea-based migrations.

  British use of indentured labourers through the East India Company carried Indo-Aryan speech even further, producing locally shaped varieties such as Fiji Hindi and Caribbean Hindustani that diverged from their source languages over generations in entirely new environments.

• 6. Romani and the Late Migration

  ›

  Romani holds a particular place in the story of Indo-Aryan dispersal. Nineteenth-century scholars Pott, writing in 1845, and Miklosich, working from 1882 to 1888, demonstrated that Romani is best understood as a New Indo-Aryan language rather than a Middle Indo-Aryan one. That finding carries a specific implication: proto-Romani speakers could not have left India significantly earlier than around AD 1000.

  The linguistic argument rests on grammar. Middle Indo-Aryan languages typically had three genders: masculine, feminine, and neuter. Romani has only two. The loss of the neuter gender is thought to have occurred during the transition to New Indo-Aryan. The neuter noun aggi meaning "fire" in Prakrit, for instance, became the feminine ag in Hindustani and jag in Romani, pointing to a shared process of gender reduction that happened after both languages remained on the subcontinent into a later period.

  Romani also maintains the Middle Indo-Aryan present-tense person concord markers and consonantal endings for nominal case, features that have disappeared from most Central Indo-Aryan languages. Its past-tense person system, meanwhile, shares an innovative pattern with Dardic languages such as Kashmiri and Shina. Linguists read this as evidence that proto-Romani speakers were originally situated in central regions of the subcontinent before moving toward the northwest. Despite all this internal linguistic evidence, no historical sources survive that document the development of Romani specifically within India.

• 7. Classification Debates

  ›

  The Inner-Outer hypothesis, first proposed by Rudolf Hoernle in 1880 and refined by George Grierson, argues that Indo-Aryan divides into a core and a periphery. Outer Indo-Aryan, which generally includes the Eastern, Southern, and sometimes Northwestern varieties, represents an older stratum that was overlaid or mixed with a newer Inner Indo-Aryan stratum. The most recent version of the hypothesis, developed by Franklin Southworth and Claus Peter Zoller, builds on a particular grammatical feature: an Outer past tense marked with -l-. Sceptics of the theory include Suniti Kumar Chatterji and Colin P. Masica.

  The Dardic languages, spoken mostly in the northwestern extremities of the subcontinent, sit at the centre of a separate dispute. George Abraham Grierson first formulated the Dardic grouping in his Linguistic Survey of India, though he did not consider it a subfamily of Indo-Aryan. Southworth has questioned whether Dardic constitutes a genuine subgroup at all, arguing that the similarities among Dardic languages may result from later convergence rather than common descent.

  Anton Kogan's 2016 lexicostatistical study, built on a 100-word Swadesh list using techniques developed by Sergei Starostin, arrived at an even stronger position: he excluded Dardic from Indo-Aryan entirely, citing low lexical similarity to Indo-Aryan at 43.5% compared with 39.3% similarity to Iranian. The modern consensus among Indo-Aryan linguists, however, still leans toward including Dardic on the basis of its morphological and grammatical features rather than vocabulary counts alone. Kogan also calculated Sinhala-Dhivehi to be the most divergent branch of Indo-Aryan, a finding tied to the fact that the Insular Indo-Aryan languages of Sri Lanka and the Maldives began diverging from the continental languages from around the 5th century BCE.