The Writing Revolution

by Amalia E Gnanadesikan

  When Brhm, Devangar’s ancestor, was first used, however, Hindi was still far in the future. Even Sanskrit written literature had yet to be born. Sanskrit was the language of sacred brahminic liturgy, and the records of secular rulers were written in the plainer, everyday Prakrits. Aoka, a northern king, naturally used the Prakrit language of the northern Indo-Aryans. In the south, however, the people spoke Dravidian languages. For the most part the Dravidian languages remained unwritten until many centuries after Aoka, with the sole exception of Tamil in the southeast. The oldest known Tamil inscriptions (written in Brhm) may date from as early as 254 bc; Tamil was to be the only written Dravidian language for some seven hundred years.

  Sanskrit itself remained unwritten (as far as we can tell from preserved inscriptions) for a century of two after the time of Aoka. Early writing in India served the same kinds of administrative purposes for which writing was usually invented and adopted in cultures around the world. Then something quite remarkable happened. Suddenly writing acquired a higher purpose: India discovered literature. The language for this new artistic use of written language was Sanskrit. The great Indian epics, the Ramayana and the Mahabharata, were written in Sanskrit. Kings hired poets to compose paeans in their praise – in Sanskrit. These poems were then included as prologues in inscriptions that recorded the acts of the kings. The prosaic details of the king’s gift of land to a temple or of tax immunity to a community of Brahmins could be written in Prakrit, but the finely crafted introductory expressions of the king’s greatness were in Sanskrit.

  With startling speed, the literary use of Sanskrit – in a form now known as Classical Sanskrit – swept the subcontinent. Kings everywhere hired Sanskrit poets and grammarians. To sponsor a work of Sanskrit grammar was a status symbol no self-respecting ruler would want to forgo. No longer confined to brahminic liturgy, Sanskrit was used by all literary people, regardless of their caste, native language, or religion. The Buddhists and Jains, whose sacred texts were originally composed in Prakrits, embraced Sanskrit enthusiastically. Classical Sanskrit was the language of literature, of kingship, of scholarly inquiry, and of education throughout the subcontinent, even in the Dravidian south. Even written Tamil was almost entirely eclipsed by Sanskrit. The Prakrits were largely confined to prosaic, documentary uses and to representing the speech of the common people (no matter what their actual language) in literary works. However, the Prakrits too were increasingly formalized and removed from anyone’s actual spoken language. Here was diglossia with a vengeance: anyone who wanted to become literate had to learn a special language.

  To write Sanskrit the Brhm script was revamped slightly, with added symbols for phonemes that occurred in Sanskrit but not in the Prakrits. Thanks to the careful phonological analysis of the Sanskrit grammarians, the adaptation of Brhm to Sanskrit was done scientifically. One of the changes was the addition of syllabic consonants to the repertoire of “vowels.” In some languages, such as Sanskrit, a consonant may sometimes take the place of a vowel in a syllable, serving as the syllable’s most resonant part. English has some syllabic consonants, such as the l of trouble, in which the written e is usually unpronounced and the l provides the resonant sound after the plosive consonant b. In Sanskrit the syllabic consonants were considered vowels, as they were doing the work normally done by vowels. So there were separate symbols for r when used as a consonant ( in Devangar) and when it was used as a “vowel” (). As a vowel, the syllabic form usually appeared as an appendage to consonants: A long version, syllabic , also occurred occasionally and received its own symbol, Careful attention to the phonology of the formalized language of which they were both students and guardians led the grammarians to realize that the words that shared one particular verb root contained a syllabic É, which was assigned the symbol . For the sake of symmetry, it was given a long partner, , although this letter was never actually used – a classic case of theory driving practice. No syllabic consonants occur in the modern languages, but the symbol for is still retained in many Indian scripts (though pronounced as ri or ru), and the symbol for remains in some.

  From about the first century BC and lasting for about a thousand years, Sanskrit was by far the chief vehicle of written language. The power of Sanskrit was such that it not only conquered the Dravidian peoples of South India but, beginning in about the fourth century AD, it also swept Southeast Asia. From Burma to Java the Southeast Asians took up writing in Sanskrit. Like the Indians, they too praised their kings in Sanskrit verses. They imported Brahmins to teach them about Hinduism, and Buddhist missionaries to teach them about Buddhism. Temples were decorated with scenes from the Ramayana, and the Thai capital city was named Ayutthaya, after Ayodhya, the city of Rama in the Ramayana. The kings of Thailand still bear the title of Rama, the present King Bhumibol Adulyadej being Rama IX.

  One of the remarkable things about the spread of Sanskrit and of Indian learned culture throughout Southeast Asia is that it happened entirely without military or political conquest. The Southeast Asians acknowledged the Chinese emperor as their remote overlord and occasionally sent him tribute. But only the Vietnamese were truly under the control of China. They adopted Chinese writing, Mahayana Buddhism (the form practiced in China), and other aspects of Chinese culture. The rest of the region, freer to choose their cultural models, looked to India, learned Sanskrit and Pali, and chose to adopt Theravada Buddhism (at the time practiced in India and Sri Lanka), Hinduism, or sometimes a combination of both, worshipping Shiva-Buddha.

  From the frontier of India in the northwest to Java in the southeast, the language of culture was Sanskrit. It was no one’s native tongue. As a vernacular language it had been dead for centuries – if in fact it had ever been one, since it may always have been a formal dialect set apart from colloquial speech. Classical Sanskrit is thus very different from Classical Latin. The original Latin classics were written by people who actually spoke Latin. Although Latin eventually died out as a spoken language and remained in use only as a learned language, medieval scholars who wrote in Latin knew they were writing in the mother tongue of Caesar and Virgil. Classical Sanskrit, by contrast, flourished only after it was “dead.” As a purely literary language it belonged to no one people and to no one piece of soil. This ethnic neutrality may have been the secret of its remarkable success. Without it, India might never have been unified, and Southeast Asia might have learned Chinese.

  Sanskrit was considered to be eternal, unchanging, and perfect. Although the language of the classics actually did vary somewhat by time and place, the uniformity in the language over thousands of miles and centuries of time is striking. A living, vernacular language would never have been so constant.

  Given such remarkable uniformity in the language, what is surprising is that the script with which it was recorded did not remain uniform at all. No one seemed to care if the Brhm used in the north diverged from that used in the south, so long as the Sanskrit it recorded was good literary Sanskrit. This lax attitude toward scripts may have been due to the respect that continued to be accorded to the oral tradition and to oral performance of literary texts. If most people experienced written works as dramatic readings, then perhaps the manner in which those works were written was not considered important. The continuing pre-eminence of the spoken word may also account for why there has never been a strong calligraphic tradition in India, despite the fact that the native scripts are inherently quite beautiful. The script was merely a vehicle for the text, and a well-educated person was expected to read many scripts. (By contrast, there are many educated Indians today who speak languages that they cannot read – they may have picked up a language rather easily in conversation, but to read it they would have to master yet another script.)

  With no one trying to keep them standardized, the regional forms of Brhm diverged from each other quite considerably. At first a simple north/south difference was visible. The northern form of Brhm eventually evolved into today’s Devangar, Bengali, Gujarati, Gurmukhi (Punjabi), Oriya, T
ibetan, and minor scripts such as Meitei-Mayak.

  The southern style spread to Southeast Asia and continued to diversify both there and in South India. Additionally, it is possible that Christian missionaries from South India brought the concept of akara writing to Ethiopia and influenced the fourth-century changes in the script there. In South India and Sri Lanka the Telugu, Kannada, Tamil, Malayalam, and Sinhala scripts are descended from the southern form of Brhm. In Southeast Asia Brhm-derived scripts are used for Khmer (of the Mon-Khmer branch of the Austro-Asiatic language family), and languages of the Tai family (Thai, Lao, and Shan) and the Tibeto-Burman branch of Sino-Tibetan (Karen and Burmese). Related forms of Indic writing also spread to the Austronesian languages of Java, Bali, Sumatra, Sulawesi, and even parts of the Philippines, where surprised Western missionaries reported universal literacy in the sixteenth century. These insular Southeast Asian scripts, however, have mostly been superseded by the Roman and Arabic alphabets.

  In Sulawesi the local language had no consonant clusters and virtually no syllable-final consonants, so conjunct consonants were done away with. When writing spread from there to the Philippines, there was no way to represent final consonants, leaving the resulting scripts closer to syllabaries than Brhm ever had been. One surviving insular script is that of the tribal Hanunóo people of Mindoro in the Philippines. The Hanunóo use their script primarily for exchanging love poetry, which they carve into bamboo or other wood with the point of a knife. The script is unusual in being generally written from bottom to top, away from the body – and, of course, for not being associated with administration, bureaucracy, or urban civilization.

  A distinctive feature of most Indic alphasyllabaries, whether northern or southern, is the head mark that letters tend to have, derived originally from serifs that resulted from using a reed pen. These serifs are now an integral and distinctive part of the Indic writing systems. In Devangar, Bengali, and Gurmukhi the head marks have become lines running across the top of each letter, which are joined within a word (as in , Devangar). Thus when writing on lined paper in these scripts one suspends the letters from the line rather than placing them on the line. In Oriya the serifs have become umbrella-like arched lines atop the letters; in Kannada and Telugu a pervasive “checkmark”; in Tamil a tendency for letters to contain a straight line across the top (see figure 10.2).

  The traditionally favored writing surfaces in India were birch bark and palm leaves (plate 7). These writing materials were readily available in India in the first millennium, in contrast with Europe and the Middle East, where parchment was rare and costly. In the north a reed pen was used, in the south a metal stylus. Since a metal stylus could easily tear the writing surface, the southern scripts developed more rounded forms, while the northern scripts remain more angular. The Oriya script, located on the border of the northern and southern areas, is closely related to the angular, northern Bengali script; but because it was written with a stylus it developed the more rounded look of a southern script.

  Figure 10.2 The Devangar, Kannada, and Tamil scripts, all descended from Brhm and following the same alphabetical order, though Tamil has reduced the number of letters.

  Not only did every region develop its own script, each region also had its own set of Brhm-derived numerals, with shapes that evolved freely as they spread throughout South and Southeast Asia. The Brhmscript is also the ultimate source of the numerals used in most of the world today, though the zero is a late (c. AD 600) addition to the system. In English we call them Arabic numerals, but the Arabs themselves call them Indian. Around AD 800 Arabs from two different regions adopted two different local forms of the Indian numerals; the forms used by the western Arabs were passed on to the Europeans. Some Indian scripts have since given up their local numeral shapes in favor of the European ones, which are now understood everywhere.

  The multitude of regional scripts in southern Asia arose mostly in the writing of a single language, Sanskrit – a perfect example of diversity retained in the face of outward conformity. Meanwhile India’s many other languages went almost entirely unwritten. Other than the Prakrits, the regional language with the longest recorded history is Tamil, lucky enough to have gotten started on the literary endeavor before Classical Sanskrit really began to dominate. The oldest examples of written Tamil are roughly as old as the Aokan Prakrit inscriptions. Tamil was given a grammar, the Tolkppiyam, around 100 BC, and a collection of poetry was produced in the following centuries. Yet even Tamil was decisively overshadowed by Sanskrit for several centuries.

  The first written Kannada (the language of Bangalore) dates from about AD 450, the first Telugu (the language of Hyderabad) from about 620, and the first Malayalam (the language of the Malabar Coast) from about 830. During the same period, several of the Southeast Asian languages first appear. For centuries after each language was first written, however, the regional languages were used only for prosaic, documentary purposes. Expressive or artistic literary work continued to be done in Sanskrit.

  Tamil was the first to shake off the Sanskritic yoke, beginning in the eighth century. Eventually the other major Dravidian languages of South India and the various languages of Southeast Asia followed suit. In the north, the switch to regional vernaculars was several centuries slower in coming. The major regional languages of the north were genetically related to Sanskrit – and even more closely related to Prakrit – so perhaps the need for written vernaculars was less strongly felt. Geography probably also played a role. The movement of peoples and their languages was freer across the broad Gangetic plain of northern India than it was in other areas with more natural boundaries. The absence of boundaries in a region encourages the development of a dialect continuum, in which only slight differences in speech are discernable from one local area to the next. In such a case it would not have been obvious which dialect to reify in writing. Even today it is not always clear where one language in North India ends and another begins; the language known as “Hindi” includes 48 recognized regional variants, not including Urdu, which is much like Hindi but uses the Perso-Arabic script. Many of these other dialects could be considered different languages – and doubtless would be if, like Urdu, they had distinct written traditions.

  When the regional languages came to be written down, they were written in whatever local version of the Brhm script was used to write Sanskrit. In the south some letters had to be added to represent phonemes used by the Dravidian languages but not Sanskrit. Sanskrit had a great number of consonants; the Dravidian languages had fewer, but nevertheless had some that Sanskrit did not. It was impossible to reuse “extra” letters from Sanskrit, because those letters were still needed for writing Sanskrit, the dominant language. In fact, Sanskrit continued to be written in a multiplicity of local scripts until the introduction of printing in the seventeenth century, since when it has been almost exclusively confined to Devangar. Alphabets all across southern Asia reflect the phonemes of Sanskrit and arrange them in the same order. The order is one carefully worked out by Sanskrit grammarians, who arranged the letters according to how they are pronounced, from the back of the mouth to the front. Extra letters, where needed, generally appear at the end of the list.

  The languages of southern Asia soaked up formal Sanskrit vocabulary like so many sponges. When they finally came to be used as literary languages, much of the vocabulary used was actually Sanskrit – a vocabulary refined over the centuries for learned and aesthetic purposes. (Similarly, most learned English vocabulary comes from the classical European languages, Latin and Greek.) The scripts, therefore, went right on writing Sanskrit words, even when embedded in local grammar.

  Sanskrit (and the northern Indo-Aryan languages generally) distinguished voiceless plosives from voiced ones (e.g. [p] from [b]), and both of those plain plosives from aspirated ones (e.g. from [ph] and [bh]). Historically, the Dravidian languages have made no distinction between plain and aspirated plosives and used voiceless and voiced plosives in separate, predictable contexts (s
o that one can say, for example, that [p] and [b] are alternative pronunciations of a single phoneme, the choice of pronunciation being determined by context). As a result, there are many letters of the alphabet that are actually pronounced alike in colloquial speech. However, people educated in Sanskrit would pronounce Sanskrit words with contrasts of voicing and aspiration duly made. Over time some of the distinctions of Sanskrit consonants have made their way into the local Dravidian languages; certain other distinctions are made only in formal contexts by educated speakers.

  The lone rebel in the story is again Tamil. Although the Tamils had a Sanskrit-based alphabet, the Grantha script, when it came to writing their own language they systematically ejected all letters that were not phonemic in Tamil. The result was the shortest Brhm-based alphabet in India. Even Sanskrit loanwords were written with the nearest equivalent letters, so that Sanskrit p, b, ph, and bh are all written p. Some Sanskrit sounds were not considered to have close-enough equivalents, so the alphabet was grudgingly given a five-letter appendix of “Grantha letters.” The use of any of these letters in a word is a dead giveaway that the word is a loanword, though these days it may be from English, not necessarily Sanskrit.

  Across India the basic way the Brhm-based scripts work is much the same: consonants and vowels receive separate treatment, sequential consonants are put together in a cluster, vowels other than the default vowel are written as appendages to the consonants except when initial, and the lack of a following vowel is indicated with a virma. Within this general pattern, variations exist. The vowels may each have their own independent letter for when they occur initially, or they may be written initially as diacritics on a “zero consonant” or “vowel support” character. The vowel-killing virma may be written below the character, as in the northern scripts, or it may be written as a mark above it, as in the southern scripts. The conjunction of consonants may be done in different ways. In Kannada and Telugu, for example, the first consonant will be written normal size and the second consonant (sometimes in a simplified shape) written smaller below it. Tamil is the rebel once again, having done away with conjunct characters entirely except for the Grantha letter which is pronounced ksa and derives from k + , but is treated as a single letter, comparable to x, [ks], in English. In all other cases, any consonant that is not followed by a vowel receives a dot on top, the Tamil version of the virma.