Skip to content

Chapter 13: Persian, Hindavi and Urdu: The Indo-Persian Sciences of Language

rekhte ke tumhīṃ ustād nahīṃ ho ghālib
kahte haiṃ agle zamāne meṃ koī mīr bhī thā

You are not the only master of Rekhta, Ghālib; they say that in an earlier age there was also one Mīr.

— Ghālib, Urdu dīvān, ghazal 36, verse 11

13.1 Introduction

For some six centuries India had a second learned culture of language beside the Sanskrit one. Persian came to northern India with the Ghaznavids and the Delhi Sultanate. From the 13th century it was a language of government, poetry, history and scholarship, and under the Mughals it was the language of administration across most of the subcontinent until the East India Company replaced it in the courts in 1837. Persian learning carried with it a set of language sciences inherited from Arabic scholarship: morphology (ṣarf) and syntax (naḥw), lexicography (lughat), prosody (ʿarūḍ), and rhetoric in its three branches of maʿānī, bayān and badīʿ. Indian scholars worked in all of them, and in one, Persian lexicography, most of the major works of the 16th to 19th centuries were compiled in India (Alam 2003).

This tradition belongs in a history of Śabda Vidyā for three reasons. It produced a large and sophisticated literature on language that surveys centred on Sanskrit usually omit. It met the Indian languages directly: Persian scholars wrote grammars of Braj and Hindavi, compiled glossaries of Indian words, and reflected on the resemblances between Persian and the languages of India. And it shaped the emergence of Hindavi, Rekhta and Urdu, the language whose 19th-century division into Hindi and Urdu is followed in Chapter 14. The literature of Urdu is treated in Volume 8, Kāvya Kalpalatā, and the wider meeting of Islamic and Indian learning in Volume 24, Samanvaya. This chapter is concerned with the scholarship of language.

13.2 Arabic Grammar in Indian Learning

In the madrasas of Muslim India, Arabic grammar held a place comparable to that of Pāṇini in Sanskrit learning. It was the gateway to every higher subject, because the Qurʾān, ḥadīth, law and theology were studied in Arabic. The curriculum that took shape in the 18th century at Farangī Maḥall in Lucknow, known as the Dars-i Niẓāmī, became the model for much of South Asia. It began with short primers of morphology and syntax, many of them written in Persian, such as the Mīzān and the Naḥw-i Mīr. It then led students through the classic Arabic grammar of Ibn al-Ḥājib, the Kāfiya, with the Persian scholar Jāmī’s commentary on it. As in the Sanskrit schools, texts were memorised first and explained afterwards, and mastery was tested by disputation.

Indian-born scholars also made major contributions to Arabic lexicography itself. Raḍī al-Dīn al-Ṣaghānī (1181–1252), born in Lahore, compiled large Arabic dictionaries in Baghdad. Muḥammad Murtaḍā al-Zabīdī (1732–1791), born at Bilgram in Awadh, settled in Cairo, where he wrote the Tāj al-ʿArūs, a vast commentary on Fīrūzābādī’s dictionary and the largest of all classical Arabic lexicons. His near-contemporary from the same town, Ghulām ʿAlī Āzād Bilgrāmī (1704–1786), wrote in Arabic as well as Persian; his Subḥat al-Marjān introduced Arabic readers to Indian figures of speech and the Indian classification of heroines (nāyikābheda) and compared them with Arabic rhetoric, a rare case of Sanskrit poetics passing into Arabic.

13.3 Persian Lexicography in India

13.3.1 Why India?

Persian lexicography flourished in India rather than in Iran for a simple reason. In Iran Persian was the mother tongue of most of its writers, who had less need of dictionaries. In India it was a learned language, acquired at school by Indian Muslims and by Hindus of the scribal and administrative classes (Kāyasthas, Khatrīs and others) who staffed the Mughal administration. These learners needed guidance on rare words, correct spelling and pronunciation, and the meanings of the words used by the classical poets. Indian lexicographers, many of them themselves second-language users of Persian, supplied it. Their position resembles that of the Sanskrit grammarians, who also described a language learned from books and teachers rather than from the household (Section 1.6).

13.3.2 From the Sultanate to the Mughals

The earliest Persian dictionaries compiled in India date from the Delhi Sultanate. The Farhang-i Qawwās of Fakhr al-Dīn Mubārakshāh Qawwās, compiled in the reign of ʿAlāʾ al-Dīn Khaljī (r. 1296–1316), arranged its words by subject, much as the Sanskrit kośas did (Chapter 4). In the 15th century the Sharafnāma of Ibrāhīm Qawwām Fārūqī, compiled in Bengal, glossed some of its Persian words in Hindavi. It is thus an early, if small, source for the vocabulary of the north Indian vernacular. A series of further dictionaries followed in the 15th and 16th centuries, each correcting and enlarging its predecessors, until the Mughal age produced the works that made India the centre of Persian lexicography.

13.3.3 The great dictionaries

The Farhang-i Jahāngīrī was begun on Akbar’s orders by Jamāl al-Dīn Ḥusayn Injū, a Shirazi nobleman at the Mughal court, and completed in 1608 for Jahāngīr. It opens with an extended introduction on the letters, orthography and grammar of Persian, so that it is a grammar as well as a dictionary, and it supports its definitions with citations from the poets. Its scale and authority made it the model for its successors.

The Burhān-i Qāṭiʿ, compiled at Hyderabad in 1651–52 by Muḥammad Ḥusayn Tabrīzī and dedicated to the Quṭb Shāhī sultan ʿAbd Allāh, owed its success above all to its arrangement. Earlier Persian dictionaries often followed the Arabic practice of grouping words by their final letter. The Burhān was arranged strictly by the order of the initial letters, which made it far easier to consult. It was also generous: it took in rare and archaic words, Zoroastrian terms, and a number of words drawn from the Dasātīr, a 17th-century work whose “ancient Persian” vocabulary was largely invented. These inclusions were later criticised severely. In 1861–62 the poet Ghālib, who prided himself on knowing Persian as the Iranians did, published a sharp attack on the dictionary (Qāṭiʿ-i Burhān), and a pamphlet war followed. Nevertheless the Burhān was printed at Calcutta in 1818 and translated into Ottoman Turkish. Together with the Jahāngīrī and the Rashīdī, it became a principal source of the European dictionaries of Persian compiled in the 19th century. Much of the Persian lexicon known to modern scholarship was thus first collected and sifted in India.

The Farhang-i Rashīdī (1653–54) of ʿAbd al-Rashīd of Thatta in Sindh marks a turn towards critical lexicography. Its author checked the entries of the Jahāngīrī and of the Farhang-i Surūrī, a dictionary compiled in Iran, against the usage of the poets. He rejected readings that he judged to be errors of copying, and he was sceptical of etymologies that rested on resemblance alone. The tradition continued into the colonial period with compact school dictionaries such as the Ghiyāth al-Lughāt of Ghiyāth al-Dīn Rāmpūrī (1826), which was reprinted many times.

13.3.4 Usage, idiom and the Hindu lexicographers

In the 18th century the emphasis shifted from rare words to idiom and usage (muḥāvara), especially the usage of the poets of the “fresh style” (tāza-gūʾī) of the 16th and 17th centuries, many of whom wrote in India. The leading works of this kind were by Hindu scholars. Lālā Tek Chand “Bahār” of Delhi, who moved in Ārzū’s circle, compiled the Bahār-i ʿAjam (1739–40), a large dictionary of Persian idioms and phrases, each illustrated by verses from the poets. Ānand Rām “Mukhliṣ”, a Khatrī official and poet, compiled a dictionary of idioms and current usage, the Mirʾāt al-Iṣṭilāḥ (mid-1740s). That some of the most authoritative Persian lexicographers of the age were Hindus is a reminder that Persian learning in India was not confined to Muslims. It was the shared professional culture of the Mughal service elite. The Bahār-i ʿAjam remains one of the richest sources for the meaning of Persian idioms.

13.4 Ārzū and the Agreement of Languages

Sirāj al-Dīn ʿAlī Khān Ārzū (1687/88–1756) was the most important Persian philologist of 18th-century India. Born in the region of Gwalior and Agra, he spent most of his career in Delhi and died in Lucknow. He was a poet, a critic, the compiler of a large biographical anthology of Persian poets (Majmaʿ al-Nafāʾis) and a teacher of poets in both Persian and Rekhta. He was also a relative and early patron of the Urdu poet Mīr Taqī Mīr. His philological works include the Sirāj al-Lughat, a critical companion to the older dictionaries; the Chirāgh-i Hidāyat, on the idioms of the later poets; the Navādir al-Alfāẓ, a corrected and enlarged edition of an earlier Hindavi–Persian glossary (Section 13.6); and the Muthmir (“Fruitful”), a treatise on the Persian language itself (Dudney 2022).

In the Muthmir Ārzū discussed what he called tavāfuq al-lisānayn, the “agreement of the two languages”. He meant the many words that Persian and the Indian vernacular (Hindī) have in common. The phenomenon is easy to illustrate: Persian mādar and Hindi mātā “mother”, Persian nām and Hindi nām “name”. Ārzū distinguished such agreement from mere borrowing in either direction, and he treated it as a matter for serious philological attention rather than coincidence. Ārzū did not propose a family tree or a common ancestral language, and his framework was not that of 19th-century comparative philology. His observations can nonetheless be compared, as a modern analogy, with the recognition a generation later that the Iranian and Indic languages are closely related (Chapter 14). There is no evidence that William Jones, whose address of 1786 drew on Persian learning, knew of the Muthmir (Dudney 2022).

Ārzū was equally modern in his view of linguistic authority. He held that languages change with time and place, and that a usage established among competent Indian writers of Persian was correct even if it differed from the speech of Iran. The view was put to the test after the Iranian poet Shaykh ʿAlī Ḥazīn settled in India in the 1730s. Ḥazīn mocked the Persian of Indian poets as faulty, and Ārzū answered him in a critique of Ḥazīn’s own verse, the Tanbīh al-Ghāfilīn (“Warning to the Heedless”). The controversy was a debate about who owns a cosmopolitan language: its native speakers, or the wider community that writes it. It has an interesting parallel in the Sanskrit grammarians’ appeal to the usage of the cultivated (śiṣṭa), not the speech of any one region, as the standard of correctness (Chapter 7). In the next century Ghālib’s attack on the Burhān reopened the question from the opposite side, in favour of Iranian authority.

13.5 Sanskrit and Persian

Learning also ran between Persian and Sanskrit. Under the Mughals many Sanskrit works were translated into Persian, among them the epics and, in 1657, the Upaniṣads by Dārā Shukōh; this history is followed in Volume 24, Samanvaya. In the other direction, a small number of Sanskrit works taught Persian to readers of Sanskrit. Audrey Truschke has traced a series of Sanskrit lexicons and grammars of Persian from the 14th to the 17th century (Truschke 2012). Two examples stand out. The Pārasīprakāśa of Kṛṣṇadāsa, composed at Akbar’s court, is a grammar of Persian in Sanskrit with a glossary, which describes Persian nouns and verbs in the categories of Sanskrit grammar, with cases as vibhakti and tenses as lakāra. A second work of the same name, by Vedāṅgarāya (1643), was written for Shāh Jahān as a glossary of the Persian and Arabic terms of astronomy and astrology. These works are the mirror image of the Persian grammars of Indian languages. Each tradition described the other in its own categories, as the Pāli and Tamil grammarians had earlier used the categories of Sanskrit (Chapters 9 and 10).

The best-known Persian account of an Indian language is Mīrzā Khān’s Tuḥfat al-Hind (“Gift from India”), written around 1675 for a Mughal prince. It is an encyclopaedia of the learning a cultivated reader needed to enjoy Hindi (Braj) poetry. It explains the Indian script and how Indian sounds, such as the retroflex and aspirated consonants, can be written in Persian letters; it gives a grammar of Braj; it treats prosody (piṅgal), rhetoric (alaṅkār), the theory of rasa and music (the last discussed in Volume 9, Saṅgīta–Nāṭya); and it ends with a glossary of Hindi words (Busch 2011). It is the first systematic grammar of Braj. It is not the first description of a modern Indo-Aryan language: Dāmodara’s Uktivyaktiprakaraṇa had explained an Old Kosali vernacular in the 12th century in the course of teaching Sanskrit, and Thomas Stephens’s grammar of Konkani had been printed in 1640 (Section 9.11). But it is among the first to treat a vernacular as a literary language with a grammar of its own. Its treatment of Braj prosody and poetics shows how much of the Sanskrit sciences of language reached Persian readers through the vernacular literatures.

13.6 Hindavi, Rekhta, Dakani and Urdu

13.6.1 The names of a language

The language that became modern Hindi and Urdu had many names before it had a standard form, and the names themselves record its history (Faruqi 2001; Rahman 2011). Hindavī or Hindī (“of Hind”) was the usual name in Persian sources from the Sultanate onwards for the vernacular of north India, as opposed to Persian. In the Deccan sultanates the same language, carried south by soldiers, officials and Sufis, developed a literature of its own from the 15th century and was called Dakanī (Dakhni); in Gujarat a related literary idiom was called Gujrī. Rekhta (“poured, mixed”) was at first the name of a style of poetry that mixed Persian and Hindavi; in the 18th century, when Delhi poets took up the ghazal in the vernacular, it became the name of the literary language itself, and Mīr and later Ghālib called their Urdu verse Rekhta. Urdu comes from the phrase zabān-i urdū-yi muʿallā, “the language of the exalted camp (or court)”, meaning Shāhjahānābād (Delhi), and is not attested as the name of the language before the last decades of the 18th century. Hindustani was the name preferred by many European writers for the common spoken language of the north.

13.6.2 Glossaries and the Persian view of Hindavi

Persian scholars first approached Hindavi through glossaries. A popular genre of rhymed vocabularies for schoolchildren, modelled on the Arabic–Persian Niṣāb al-Ṣibyān, produced Persian–Hindavi versions. The best known is the Khāliq Bārī, traditionally attributed to Amīr Khusrau (1253–1325), though most scholars consider it a later work. In the late 17th century ʿAbd al-Wāsiʿ Hānsawī compiled the Gharāʾib al-Lughāt, a glossary of Hindavi words with their Persian equivalents. Ārzū’s corrected version of it, the Navādir al-Alfāẓ, adds critical comments on pronunciation and usage. Together they are among the most valuable records of the vocabulary of north Indian speech before the 19th century. Walter Hakala has shown how these works helped define what counted as a language and as a correct word in the multilingual world of Mughal India (Hakala 2016).

13.6.3 European grammars and Gilchrist

European interest in Hindustani began with the trading companies. The Dutch East India Company’s agent Joan Josua Ketelaar wrote a grammar of Hindustani in Dutch around 1698, of which a Latin version was printed at Leiden in 1743; the Halle missionary Benjamin Schultze published a Grammatica Hindostanica in 1745; and the British officer George Hadley published grammatical remarks on the “vulgar dialect of the Indostan language” in 1772. The decisive figure was John Borthwick Gilchrist (1759–1841), a Scottish surgeon. His Dictionary, English and Hindoostanee (1787–90) and Grammar of the Hindoostanee Language (1796) made him the leading European authority, and in 1800 he became the first head of the Hindustani department of the College of Fort William in Calcutta. Gilchrist distinguished several registers of the language: a Persianised “High Court” style, a “Middle” or genuine Hindustani style that he recommended to British officers, and a “Hinduwee” style closer to rural speech and Sanskrit. He insisted on a Roman transcription of his own design. The prose texts he commissioned from his Indian staff at Fort William, in what came to be called Urdu and Hindi, and the institutional consequences of separating them, are discussed in Chapter 14.

13.6.4 Inshā’s Daryā-yi Laṭāfat

The first comprehensive description of Urdu by Indian authors was written, characteristically, in Persian. The Daryā-yi Laṭāfat (“Sea of Elegance”, 1807–08) is by the poet Inshā Allāh Khān “Inshā” (1752–1817) of Lucknow, with a second part by his colleague Mirzā Muḥammad Ḥasan Qatīl on prosody and rhetoric. It covers the sounds and grammar of Urdu. More unusually, it describes the varieties of the language as spoken by different groups: the people of the various quarters of Delhi, the speech of women, and the usage of Hindus and Muslims, Punjabis and easterners. Inshā’s standard of eloquence (faṣāḥat) was the speech of the cultivated Delhi elite, and he judged other varieties against it. The standard tied linguistic correctness to social rank, and the book’s portraits of other people’s speech are often condescending. As a record of linguistic variation around 1800, however, it has no parallel. Inshā was also a virtuoso of registers. In his Rānī Ketakī kī Kahānī he set himself to write a story with no Persian or Arabic words at all. It is an early sign of the consciousness of vocabulary that was later to separate Hindi from Urdu.

13.7 Assessment

The Indo-Persian sciences of language were scholarship by and for second-language users of a cosmopolitan language. That is why they were so strong in lexicography and usage, and why questions of authority, of who decides what is correct, were so central to them. In this they resemble the Sanskrit tradition more than is usually recognised. They also connect that tradition with the modern period. The Persian dictionaries of India became sources of European Persian scholarship. Ārzū’s reflections on the agreement of Persian and Hindi show that the resemblances between Iranian and Indian words were a subject of philological attention in India a generation before Jones’s address, though within a framework of their own that did not include a common ancestor. And the Persian descriptions of Braj, Hindavi and Urdu, from Mīrzā Khān to Inshā, are among the earliest grammars of the north Indian vernaculars; the oldest of them precede the European grammars of Hindustani that are usually treated as the beginning of their study.

Further Reading

  • Alam, Muzaffar. 2003. “The Culture and Politics of Persian in Precolonial Hindustan.” In Literary Cultures in History: Reconstructions from South Asia, edited by Sheldon Pollock. University of California Press.. — On Persian in precolonial north India.

  • Dudney, Arthur. 2022. India in the Persian World of Letters: Khān-i Ārzū Among the Eighteenth-Century Philologists. Oxford University Press.. — On Ārzū, his dictionaries and his philology.

  • Hakala, Walter N. 2016. Negotiating Languages: Urdu, Hindi, and the Definition of Modern South Asia. Columbia University Press.. — On Persian–Hindavi lexicography and the definition of Urdu and Hindi.

  • Truschke, Audrey. 2012. “Defining the Other: An Intellectual History of Sanskrit Lexicons and Grammars of Persian.” Journal of Indian Philosophy 40 (6): 635–68.. — On Sanskrit lexicons and grammars of Persian.

  • Busch, Allison. 2011. Poetry of Kings: The Classical Hindi Literature of Mughal India. Oxford University Press.. — On Braj poetry at the Mughal courts and the world of Mīrzā Khān.