Skip to content

Chapter 12: The Manuscript Tradition: Knowledge on Leaf, Bark and Paper

12.1 Introduction

Oral transmission remained the ideal for the Veda, but for almost everything else — philosophy, grammar, medicine, astronomy, mathematics, law, literature, drama, music, ritual manuals, devotional poetry, administrative records — India relied increasingly on writing. Over two millennia scribes produced manuscripts in dozens of scripts and languages on palm leaf, birch bark, paper, cloth and other materials. India today holds one of the largest surviving bodies of manuscripts in the world, and most of the texts discussed in this volume are known to us only through them. This chapter describes the materials, scripts and practices of the manuscript tradition, its institutions, and the efforts to preserve and study it. For the history of Indian scripts see also Volume 10 of this series (Salomon 1998).

12.2 Writing Materials

12.2.1 Palm Leaf

The most widely used material, especially in the south and east, was the dried leaf of two palms: the palmyra (Borassus flabellifer) and the talipot (Corypha umbraculifera). Leaves were cut into long, narrow strips, boiled or soaked and dried to make them supple and durable, smoothed, and pierced with one or two holes through which a cord was passed to hold the leaves together between wooden boards.

Two methods of writing were used. In the south and in Odisha, letters were incised into the leaf with a metal stylus and the grooves blackened by rubbing in lamp-black or a similar pigment mixed with oil. In the north and east, letters were written on the surface with a pen and ink. It is often suggested that the rounded forms of scripts such as Odia and Malayalam developed because straight horizontal strokes would split an incised palm leaf along its grain.

Ink for writing on palm leaf and paper was usually made from lamp-black or charcoal mixed with gum and water, sometimes with other ingredients to improve its adherence and colour; red, yellow and other coloured inks and pigments were used for headings, corrections and illustrations, and gold and silver inks for luxury manuscripts. Pens were made of reed or bamboo. Many manuscripts carry marks and corrections showing that they were read, checked and used in teaching, not merely stored.

12.2.2 Birch Bark

In Kashmir and the north-west, the inner bark of the Himalayan birch (bhūrja, Betula utilis) was the main writing material, written on with ink. The oldest surviving Indian manuscripts are birch-bark scrolls in the Gāndhārī language and Kharoṣṭhī script, containing Buddhist texts, found in Afghanistan and Pakistan and dated to between the first century BCE and the third century CE. Other famous birch-bark manuscripts include the Gilgit manuscripts (5th–7th centuries), which preserve many Buddhist texts in Sanskrit; the Bower Manuscript, a medical collection found at Kucha in Chinese Central Asia and usually dated to about the sixth century; and the Bakhshālī Manuscript, a mathematical text whose date is disputed (radiocarbon tests published in 2017 gave widely separated dates for different leaves, which scholars have interpreted in different ways).

12.2.3 Paper

Paper, introduced through the Islamic world, came into use in India from about the eleventh to the thirteenth centuries, first in the north-west and in Gujarat, where some of the earliest dated paper manuscripts are preserved in Jain libraries. Over the following centuries it gradually replaced birch bark in the north and became common everywhere, though palm leaf remained in use in the south and east into the nineteenth century.

12.2.4 Other Materials

Palm leaf, birch bark and paper were not the only surfaces on which Indians wrote; Table 12.1 lists the others.

Table 12.1. Other writing materials used in India.
Material Use
Stone Inscriptions on rock, pillars, temple walls; the edicts of Aśoka (3rd century BCE) are the earliest securely dated deciphered writing in India
Copper plates Royal land grants and charters; many thousands survive and are a principal source for dated history
Cloth Jain cosmological and ritual diagrams (paṭa); some texts and paintings
Wooden boards Covers for palm-leaf manuscripts, often painted; occasionally for writing
Agarwood bark In Assam, the prepared bark of the agar tree (sāñcipāt) was used for manuscripts

12.3 Scripts

The writing systems of South Asia, apart from those derived from Arabic, descend from Brāhmī, first attested in Aśoka’s inscriptions, and the related Kharoṣṭhī script of the north-west, which was written from right to left and fell out of use in the early centuries CE. Brāhmī developed regional forms that by the end of the first millennium had become distinct scripts (Salomon 1998); Table 12.2 lists some of them.

Table 12.2. Some scripts of Indian manuscripts.
Script Region Use
Kharoṣṭhī North-west Gāndhārī Buddhist texts; inscriptions (to c. 3rd–4th c.)
Śāradā Kashmir Sanskrit texts, especially on birch bark
Nāgarī (Devanāgarī) North and west Sanskrit, Prakrit and many regional languages
Bengali, Odia, Maithili, Newari East and Nepal Sanskrit and regional literatures
Grantha Tamil country Sanskrit texts
Tamil, Vaṭṭeḻuttu Tamil country, Kerala Tamil literature and records
Malayalam Kerala Malayalam and Sanskrit
Telugu–Kannada Deccan Telugu, Kannada and Sanskrit
Nandināgarī Karnataka Sanskrit texts, especially of the Mādhva school
Siddham Exported to East Asia Sanskrit Buddhist texts in China and Japan
Arabic and Persian scripts Throughout, chiefly from the 13th c. Arabic, Persian, later Urdu and other languages

Sanskrit, in particular, was never tied to a single script: the same text might be copied in Śāradā in Kashmir, Nāgarī in Banaras, Bengali in Navadvīpa and Grantha in Kāñcī. A scholar who wished to read widely had to be able to read several scripts.

12.4 Scribes and Copying

Copying was done by professional scribes (lekhaka, lipikara), by communities of record-keepers such as the Kāyasthas, by monks and nuns, and by scholars making copies for their own use. The Arthaśāstra describes the qualifications of a royal scribe and the faults of a badly drafted document. In Jain and Buddhist communities copying a text, or paying for it to be copied, was a meritorious act, and many manuscripts record the name of the donor and the purpose of the gift.

In Mahāyāna Buddhism the book itself became an object of veneration. Several Mahāyāna sūtras, among them the Prajñāpāramitā texts and the Lotus Sūtra, declare that copying, reciting and worshipping the sūtra brings immense merit, and that the place where it is kept should be honoured like a shrine. The historian Gregory Schopen has argued that this “cult of the book” was an important element in the early development of the Mahāyāna (Schopen 1975). It helps to explain both the care with which such texts were copied and decorated and the survival of so many copies of them.

Manuscripts frequently end with a colophon recording the title of the work, the name of the copyist, the place and date of copying (in one of several eras, such as the Vikrama or Śaka era), and sometimes the patron. Scribes often added a verse asking readers to look after the manuscript or apologising for errors and describing the bodily strain of copying. Colophons are a principal source for the history of texts and of the people who copied them.

Reading a colophon requires knowing the eras in which dates were expressed. Among the most common are the Vikrama era, which begins in 57–56 BCE and was widely used in north and west India; the Śaka era, beginning in 78 CE, used especially in the south and in astronomical works; the Kollam era of Kerala, beginning in 825 CE; the Nepal era, beginning in 879 CE; and, in Islamic contexts, the Hijrī era, beginning in 622 CE. Dates were often given with the lunar month, fortnight, lunar day and weekday, which allow modern scholars to verify them. Manuscripts also use characteristic abbreviations, correction marks and numbering systems, and the study of these features (palaeography and codicology) is a specialised discipline.

Every copying introduced errors, and copies of copies multiplied them. Indian scholars were aware of this, and commentators frequently discuss variant readings (pāṭhabheda). The modern discipline of textual criticism, applied to Indian texts from the nineteenth century, aims to reconstruct the history of a text from its manuscripts; its greatest monument is the critical edition of the Mahābhārata prepared at the Bhandarkar Oriental Research Institute, Pune, between 1919 and 1966, based on the collation of hundreds of manuscripts.

12.5 Illustrated Manuscripts

Manuscripts were also works of art (Losty 1982). Palm-leaf manuscripts of the Prajñāpāramitā and other Buddhist texts from the Pāla period in eastern India (11th–12th centuries) contain small paintings of deities and scenes from the life of the Buddha, and their painted wooden covers survive in Nepal and Tibet. In western India, Jain manuscripts of the Kalpasūtra and other texts, on palm leaf and later on paper, were illustrated in a distinctive style, with the mothers’ dreams of the Jinas (Chapter 7) a favourite subject. Odia palm-leaf manuscripts were illustrated by incision. Under the Mughals, Persian manuscripts — including the Razmnāma, the Persian translation of the Mahābhārata made for Akbar — were illustrated in the imperial workshops. Scientific manuscripts contained diagrams, tables and astronomical figures.

12.6 Libraries

  • Monastic libraries. The Buddhist monasteries held large collections (Chapter 11); many texts that perished in India survive in manuscripts carried to Nepal and Tibet or in Tibetan and Chinese translations.

  • Jain libraries. The Jain bhaṇḍāras of western India, such as those of Patan and Jaisalmer, kept in temple vaults and cared for by the community, are among the most important collections of medieval Indian manuscripts, Jain and non-Jain alike. The Jain festival of Jñānapañcamī, on which manuscripts are ritually honoured, aired and re-wrapped, helped to preserve them.

  • The Jain libraries also illustrate how manuscript collections were built. Monks and wealthy laymen commissioned copies of texts as acts of merit, often in large numbers, and the libraries were enlarged by generations of monastic leaders; the great library at Jaisalmer, for example, is associated with the fifteenth-century monk Jinabhadrasūri. Because the vaults were in the dry desert climate of Rajasthan and Gujarat, and because the community regarded the manuscripts as sacred objects to be protected, they preserve some of the oldest dated paper and palm-leaf manuscripts in India, including copies of Buddhist, Brahmanical and literary works that were not themselves Jain.

  • Temple and monastic (maṭha) libraries. Many temples and maṭhas in the south kept manuscripts, some of which survive.

  • Royal libraries. Kings collected manuscripts and patronised copying. The Sarasvati Mahal Library at Thanjavur, begun under the Nāyaka rulers and greatly enlarged by the Maratha ruler Serfoji II (r. 1798–1832), is the best-known surviving example; it holds manuscripts in Sanskrit, Tamil, Marathi, Telugu and other languages.

  • Family collections. Scholarly families kept manuscripts for generations. Large numbers remain in private hands.

12.7 Preservation

Heat, humidity, insects and fungi destroy manuscripts quickly in most of India. The principal traditional means of preservation was therefore not conservation but re-copying: a text survived because it was copied before its manuscript decayed, and a text that ceased to be copied was lost. Manuscripts were also protected by wrapping them in cloth, storing them between boards and in boxes, airing them periodically, and keeping them with substances believed to repel insects; the details varied from region to region and are known mainly from observation of surviving practice.

The dependence on re-copying explains a striking feature of Indian manuscript collections: most surviving manuscripts are a few centuries old even when the texts they contain are much older. Survival depended on continued interest; texts of schools that died out — many Buddhist works, for example — were lost in India even if they survived abroad.

12.8 The Modern History of Indian Manuscripts

From the late eighteenth century, European scholars and colonial officials began to collect Indian manuscripts. The Asiatic Society, founded in Calcutta in 1784, built up a major collection; Colin Mackenzie, surveyor-general of Madras, assembled a vast collection of manuscripts and records from the Deccan and the south; and the governments of the presidencies commissioned searches for Sanskrit manuscripts in the later nineteenth century. Many manuscripts were taken to European libraries; many others were catalogued and deposited in Indian institutions such as the Government Oriental Manuscripts Library in Madras (now Chennai), the Oriental Research Institute in Mysore, the Adyar Library in Madras and the Bhandarkar Oriental Research Institute in Pune.

12.8.1 Catalogues

The first step in using manuscripts is to know where they are. From the nineteenth century, catalogues of individual collections were published, and the German Indologist Theodor Aufrecht compiled from them the Catalogus Catalogorum (1891–1903), an alphabetical register of Sanskrit works and authors with the collections in which their manuscripts were recorded. A New Catalogus Catalogorum, published by the University of Madras from 1949 and still in progress, aims to update and greatly expand it. Similar efforts exist for Tamil, Persian and other manuscript traditions. These catalogues reveal both the enormous number of works that survive and the many that are known only by name.

12.8.2 From Manuscript to Print

Printing with movable type was introduced to India by Portuguese missionaries at Goa in 1556, but for a long time it was used mainly for Christian works. In Bengal, Nathaniel Brassey Halhed’s grammar of Bengali (1778) was printed with Bengali types cut under the direction of Charles Wilkins, and the Baptist mission press at Serampore, founded in 1800, printed grammars, dictionaries and translations in many Indian languages. From the 1820s Indian presses multiplied, and by the later nineteenth century printed books — textbooks, religious texts, newspapers, novels — were transforming reading and education. Printing did not end manuscript production at once, and many traditional scholars continued to prefer manuscripts for decades; but the printed book, together with the school system and the examination, gradually replaced the manuscript and the teacher’s recitation as the primary vehicle of learning (Kumar 1991).

Printing also changed the status of texts. A manuscript was copied for a particular person or library and might differ from every other copy; a printed edition fixed a single text and made it available to anyone who could read. The editions of Sanskrit, Pali, Tamil and Persian texts produced by Indian and European scholars from the nineteenth century onwards made classical learning accessible far beyond the communities that had traditionally controlled it — a change as significant, in its way, as the original turn from oral to written transmission.

12.8.3 Recovered Texts

Some of the most important works discussed in this series were recovered from manuscripts only in the last hundred and fifty years. The Tamil scholar U. V. Swaminatha Iyer (1855–1942) spent decades searching monasteries and private homes in the Tamil country for palm-leaf manuscripts, and published from them the Cīvakacintāmaṇi (1887), the Cilappatikāram (1892) and many of the Caṅkam anthologies, which had been almost forgotten; much of classical Tamil literature as it is now read depends on his editions. In 1905 R. Shamasastry of the Oriental Research Institute in Mysore identified a palm-leaf manuscript of the Arthaśāstra, which had been known only from quotations; his edition appeared in 1909. In 1909–10 T. Gaṇapati Śāstrī discovered in Kerala palm-leaf manuscripts of thirteen plays that he attributed to the early dramatist Bhāsa. These discoveries depended on manuscripts that had survived in private hands and on the scholars who knew how to read them.

12.8.4 Indian Texts Recovered Abroad

Many Sanskrit works lost in India survived in the Himalayan regions. In Nepal, Buddhist and Śaiva manuscripts were preserved in monasteries and private collections, and the British resident Brian Houghton Hodgson sent large numbers of Sanskrit Buddhist manuscripts from Kathmandu to libraries in Calcutta, London and Paris in the 1820s–40s. In Tibet, Sanskrit manuscripts brought by Indian teachers and Tibetan translators were kept in monasteries for centuries. Between 1929 and 1938 the Indian scholar Rāhula Sāṅkṛtyāyana made four journeys to Tibet, photographing and copying Sanskrit manuscripts, among them works of the Buddhist logicians Dharmakīrti and Prajñākaragupta that had been known only in Tibetan translation; his materials became the basis of many later editions. The history of these texts shows how much of India’s intellectual heritage depended on transmission beyond India’s borders.

12.8.5 Arabic and Persian Manuscripts

India also holds very large collections of Arabic and Persian manuscripts, reflecting six centuries of Indo-Islamic learning and administration (Chapter 11): histories, poetry, Sufi literature, theology and law, medicine and the sciences, and Persian translations of Sanskrit works. Among the best-known collections are the Khuda Bakhsh Oriental Library in Patna, opened to the public in 1891, and the Raza Library in Rampur, built up by the Nawabs of Rampur. These collections are an essential, and often neglected, part of India’s manuscript heritage.

12.8.6 The National Mission

In 2003 the Government of India established the National Mission for Manuscripts to survey, document, conserve and digitise manuscripts throughout the country. The Mission has estimated that India holds about ten million manuscripts, in a great many languages and scripts, and has documented several million of them in its database; these figures are estimates and are revised as the survey proceeds. A large proportion of the manuscripts in India have never been catalogued in detail, let alone studied, and much remains in private collections in poor conditions.

The study of manuscripts — cataloguing, editing, palaeography — is an essential part of research on Indian knowledge traditions, but it is a specialised skill that relatively few scholars now possess. Claims that uncatalogued collections contain “lost” works of famous authors or unknown scientific discoveries should be treated with caution until the manuscripts in question have been examined and published.

Further Reading

  • Salomon (1998): the standard guide to Indian scripts and inscriptions.

  • Losty (1982): the art of the illustrated manuscript in India.

  • Scharfe (2002): on writing and education.