What is a book? Physically, it could be said, it is a collection of pieces of paper of the same size between covers and held together by glue and string. In another sense, it is the intellectual content conveyed by the words which are written or printed on the pages. All cultures are in agreement with the latter statement; comparatively few would agree with the former. Two thousand years ago in the Mediterranean world a book was physically a long roll of papyrus or of parchment. At the same time in China it was a collection of thin strips of bamboo or pieces of silk.
In India, the earliest concept of the book was as a collection of leaves or sheets of bark strung together between covers by a cord. The Indians would have viewed with horror both the slaughtering of young animals for their skin and the writing of sacred texts on such material, while paper, which was invented in China in the 1st century, did not come into general use in northern India before the 13th century, at about the same time as in Europe, in each case the Muslim world being the intermediary. In southern India, however, palm leaves continued in general use until the 19th century as the normal writing material.
The first references to writing in India, found in the earliest layers of the Pali Buddhist Canon of about the 5th century BC, speak of various types of material used for writing, such as leaves (paṇṇa), wood (phalaka, or boards, and śalākā, or bamboo chips or slips), and metals. The type of leaves is unspecified, but there is no reason to believe that it is as yet actually the usual writing palm of ancient India, the talipot (Corypha umbraculifera), for the latter is indigenous only to the extreme south of the peninsula, of which the early Buddhist texts had no knowledge. In fact any kind of suitable leaf was probably made use of, as leaves of the plantain and śāla trees were used in village schools until recent times.
Since much of what is known of the social history and material culture of ancient India is deduced from the obiter dicta of authors actually writing about something else, any argument 'from silence' must be used with extreme caution. Nonetheless, there is, as Rhys David long ago pointed out, an absolute silence about books (as physical objects, that is) in the Buddhist Canon, despite long inventories of what monks are and are not permitted to own, which argues that literary or religious texts were not committed to writing, while the repeated assertion that suttas (the Buddha's discourses) could be lost through a monk's having no disciple to teach them to, argues very strongly that the mere possibility of writing down the Buddhist sacred texts could not be entertained. This is not only the Indian aversion to the written as opposed to the oral tradition, but the very real problem that no writing material known in the 6th and 5th centuries BC was usable for writing connected literary or scriptural texts, as opposed to records, letters, or accounts. It is clear therefore that the use of the talipot must have been unknown to the north of India at this time, and seems still to have been unknown by the late 4th century BC, since it is not included with the writing materials (bark and cloth) noted by the Greek companions of Alexander. It can only have been with the expansion of the Mauryan empire into the south of India in the 3rd century BC that the talipot could have become known to the northern Indians, and its possibilities exploited for the writing of literary and religious texts.