Showing posts with label Scripts. Show all posts
Showing posts with label Scripts. Show all posts

24 July 2009

Indo-European Writing


brhami scipt from British Museum

Aśoka (6th) Pillar fragment
Brāhmī inscription
formerly at Meerut. Mid-3rd century BCE.
(?u)pagamane sememokhyamate
(bhi)sitename iyaṃdhaṃmali(pi) li*


© British Museum (my photo)
When I recently blogged about Indo-European languages one of my friends asked what is quite an obvious question: what about writing? So I'm going to outline the development of writing as it is relevant to Buddhist India. India is very linguistically diverse - there are three distinct language families, including Indo-European, plus a number of languages not related to any known language - known as 'isolates'.

Most modern scripts can trace their origins to Mesopotamia. A surprising number trace their roots to the writing developed in India - in fact almost every country that saw Buddhist missionary activity, excepting China, has seen some influence from the India scripts on their writing systems. That said writing came relatively late to India, and even then was not used for religious texts for many centuries.

The earliest evidence of writing anywhere in the world is from around 3500 BCE in present day Iraq. The Sumerians left caches of clay tablets at sites along the Tigris and Euphrates Rivers. This earliest writing, similar in some ways to the earliest Chinese writing (ca 1400 BCE), involves pictures which represent concepts. However it was with the invention of cuneiform that something like writing as we know began. The earliest examples of cuneiform combine signs for concepts with signs for sounds (as do Egyptian hieroglyphs). A variety of cuneiform continued to be used in the middle-east until the last century before the common era. However in the same region speakers of Semitic languages began to represent their language using only signs for sounds - i.e. to use a true alphabet - around 2000 BCE.

The Achaemenid empire (ca 550 - 350 BCE) founded in Persia adopted a form of Semitic writing, often called Aramaic after the language it encoded, for administrative purposes. It represented a significant improvement on cuneiform - it could be written (and read) more quickly and easily, and on a much greater variety of materials. Now it so happens that the Achaemenids invaded India and controlled, or at least had a powerful influence, up to the western bank of the Indus River - about half of what is now Pakistan. In fact the Buddha lived during this period and there is some evidence of Persian influence in the Pāli Canon - which I have discussed previously in this blog. The Achaemenids were toppled by Alexander of Macedonia, and in India at least the power vacuum was filled by the Mauryan Dynasty of which Asoka is the stand out figure.

Certainly by the time that King Asoka ruled India (mid 3rd century BCE) there was a well developed form of writing: the Brāhmī Lipi or Writing of God which he used on his rock pillars and edicts. Brāhmī is said to show signs of influence from Aramaic. Another script, Kharoṣṭhī, was less widely used in India proper, but was the main script in Gandhāra and parts of central Asia for some centuries. We can say with some confidence that Kharoṣṭhī was based on Aramaic as it retained many features of the Semitic script. It is possible that Kharoṣṭhī influenced the development of Brāhmī , but it is difficult to say because the earliest known examples of Brāhmī are already a fully fledged script and the direction of development subsequently is determined by the nature of Indian languages.

We can say that all forms of written language in India, as well as Tibet, Burma, Sri Lanka, Thailand, Laos and Cambodia are descendants of Asoka's Brāhmī. In India there was a very distinct north/south divide which I think was caused by writing materials, but may have been influence by the geographical spread of Indo-European languages (north) and Dravidian languages (south). The North favoured birch bark and wood to write on, while in the South the material of choice was the leaf of the talipot palm. The palm leaf is ribbed which lead to the development of the more rounded shapes of Southern writing (which then influenced Sinhala, Burmese etc). Note that technically Sinhala is a North Indian language, part of the Indo-European family, and not related to Tamil or other Dravidian languages, but it is written in a southern style script and palm leaves were the medium of choice until relatively recently. Similarly Burmese, though written with a Brāhmī derived script is part of the Tibeto-Burman family which has a relationship with the plethora of Chinese languages.

Brāhmī underwent continuous development in the North and diversified into local geographical variants. But because dynasties based in the North-east - particularly the Mauryans and then the Guptas (3rd - 6th centuries) - were dominant they tended to influence the development of writing more. The Brāhmī variant commonly used by the Guptas was by that time a distinct script. The Nalanda University was in the Gupta heartland and the Gupta script was important in the spread Buddhist texts. Later it would give rise to the Siddhamatṛkā or just Siddhaṃ script which was commonly used to write the Tantras. It is still in use today for writing mantras in China and Japan. The Gupta script, or something very like it was used as the basis for Tibetan writing, which also continued to develop independently and diversify. Siddhaṃ evolved into Devanāgarī, which is the most common script for Sanskrit in the present - though it is written using the Tamil script in South India for instance.

There are several ways of represent sounds using signs. The English alphabet attempts to convey individual sounds that combine into syllables or phonemes. This is a very efficient way of representing spoken sounds - with just 26 letters we manage to convey the 25 single consonant sounds, and 23 vowel sounds that are used in 'standard' English (as represented by the Oxford English Dictionary) and an extensive repertoire of combination of them in words and sentences. Another method is to represent speech as a series of syllables. One can either represent all possible syllables by individual signs which is only practicable if there are a small number of syllables - the Japanese Kana alphabets are good examples of this approach. An intermediate option is available where consonants and initial vowels have distinct signs and these are modified to show medial and final vowels. The latter is typical of Indian writing.

There are several distinct features shared by all Indic scripts which I will demonstrate using Devanāgarī. Initial vowels have distinct signs - 14 in Sanskrit - but medial and final vowels, and the absence of a vowel (which happens at the end of words and in conjuncts) are indicated with diacritic marks. Consonants are assumed to be followed by the short 'a' vowel unless otherwise specified. क is ka not k. Examples of vowel diacritics are: kā ke kai ki kī ko kau kṛ = का के कै कि की को कौ कृ. Vowels may also be absent, nasalised or aspirated (the technical terms being virāma, anusvāra, visarga), so g, gaṃ, and gaḥ = ग् गं गः . Indic, like English, allows for a variety of combinations of consonants without intervening vowels. These are either written as a vertical stack as in ṣ + ṭha > ṣṭha = ष् + ठ > ष्ठ; or as a horizontal combination with the initial consonant as a "ligature": t + pa > tpa = त् + प > त्प. As many as four consonants can be combined in this way e.g. strya स्त्र्य. Special variations occur with 'r' viz pra = प् + र > प्र, and rta = र् + त > र्त; ś can also undergo a special change e.g. śva = श् + व > श्व; cf śya = श्य; and some conjuncts have distinct signs jña ज्ञ and kṣa क्ष. Writing Sanskrit this way is considerably more complex than writing English.

Buddhist texts in Indic languages are preserved in a plethora of scripts: Brāhmī in many varieties, the Gupta script again in many varieties, Siddhaṃ with some variation over time (and major variants in China and Japan), Kharoṣṭḥī, Devanāgarī, several Tibetan scripts, Sinhala, Burmese, Thai etc. Each of these scripts records texts in an equally wide variety of Indian languages and dialects; and of course many texts are now known only in translation in Chinese, Tibetan or any of a dozen other languages - making the history of Buddhist texts quite complex. All this is remarkable given centuries of resistance to the use of writing for the purpose of writing sacred texts in the early days of Buddhism.

A scholar of Buddhism must often know several languages and associated scripts in order to read the relevant manuscripts. The standard of handwriting in ancient times was often quite poor, and the attention to copying of texts wavered causing scribal errors, which means that no two manuscripts are ever the same. Deciphering such manuscripts requires imagination and powers of deduction as well language skills. As such we owe a great deal to those people down the ages with the aptitude and motivation to take on the difficult task of learning these, often dead, languages and scripts, in order to study and translate the texts for us. Having dipped into their world I am in awe of them.

Note

* The bottom line of the Brāhmī inscription pictured is a fragment of a longer stock phrase:
saḍuvīsativābhisitena me iyaṃ dhaṃma lipi likhāpita
when I [Asoka] had been consecrated twenty-six years I ordered this inscription of the dhamma to be engraved
- c.f. the 1st pillar in A. L. Basham The Wonder that was India, 1967, 2001, p.395. There is now some doubt as to whether Asoka meant the Buddhadhamma or his own dhamma [see Aśoka, Pāli and some red herrings].

01 August 2008

Which script?

Indian languages, and in particular Sanskrit, may be unique in that there is a distinct separation between the language and the scripts used to write it. Over the centuries a number of scripts and variation have been particularly important in preserving Buddhist texts. Buddhists were, like their Brahmin counterparts, initially reluctant to employ writing for the preservation of texts that were composed and transmitted orally. But eventually in several places, notably Sri Lanka and Gāndhāra, the sūtras were committed to writing. Here are the broad outlines of the development of Indian writing as it relates to Buddhist texts in roughly chronological order.


The first script to be used for writing Indian languages was Kharoṣṭhī beginning in the 3rd or 4th century BCE. It was used for several hundred years to write the local dialects, but also Sanskrit. Kharoṣṭhī was used in Gāndhāra in North West India (what is now the Pakistan shading into Afghanistan - i.e. Taliban country) and in central Asia. It shares many features with the Aramaic used by Persian administrators at the time when they were influential in that region, and most scholars accept that Aramaic was probably the model for Kharoṣṭhī. It was both carved in rock and written with pen and ink on birch bark. Quite a number of early Buddhist texts survive in Sanskrit and in the Gāndhārī Prakrit (roughly the vernacular dialect of Gāndhāra). A Gāndhārī version of the Dhammapada survives for instance, and several other texts which parallel the Pāli but with interesting mostly minor differences. These texts have helped to flesh out our knowledge of the different early schools of Buddhism.


Brahmī was the first truly indigenous Indian script. The name means simply "God" - Brahmā having been adopted as a creator god by this point. It was a definite improvement on Kharoṣṭhī in having individual signs for initial vowels, and greater variation between characters which made it more readable. My guess however is that it was designed for carving rather than writing per se. The signs are simple, geometric and quite linear. Aśoka (mid to late 3rd century BCE) used this script for most of his rock edicts - a few were in Kharoṣṭhī or Aramaic, and one was in Greek reflecting the local usage in the far West of his imperium. Probably most Buddhist texts were written in this script in Eastern and Central India until the 3rd century CE.

An early south Indian variety of Brahmī became very important. The Sinhala script emerges around the 2nd or 3rd centuries BCE and was used to commit the Pāli Canon to writing. It shows the rounded letter forms characteristic of writing designed for palm leaves with their pronounced grain that could be easily punctured.

In the North the Gupta script (so-called because it emerged during the Gupta Era, ca 3rd - 6th centuries) also derives from Brahmī. It shows the influence of pen and ink writing (probably still on birch bark). Versions of the Gupta script were used throughout North Indian and central Asia to transmit Buddhist texts. Many of the early Buddhist texts going to China would have used this script or one of its Central Asian variants. Some use of decorative writing - what we might call calligraphy - began to appear at this time.


Siddhaṃ was initially a word written in the top left hand corner of any piece of writing in the Gupta period, meaning perfection or accomplishment. However as the script changed and the letters became more elaborate Siddhaṃ (or more fully siddhamātṛka) began to be the name of the script. By the collapse of the Gupta Empire (under attack by Huns) in the 6th century the script had become distinct. However it continued to undergo development for several centuries. Two forms are commonly seen nowadays that I call "Chinese brush style", and "Japanese pen style". The latter appears to be a further refinement of the Indian script, while the former is influenced in it's form by the demands of writing with a brush. This script was important in the East as the medium of the early Tantric texts. What's more Tantric Buddhism initially inherited Vedic injunctions to preserved accurate pronunciation of letters and so the Indic script was preserved especially in the case of mantras - they can still be seen in the modern Taisho version of the Chinese Tripitaka for instance. Siddhaṃ calligraphy was elevated to a fine art by the Japanese. While Kūkai introduced Siddhaṃ to Japan and produced some fine works, the modern popularity derives from an 18th century revival. Siddhaṃ is mainly used for writing mantras and bījas these days, although shakyo or sutra copying has not completely died out - the Heart Sūtra being a favourite text.

An early variant of Siddhaṃ formed the model for the Tibetan Uchen (dbu-can) script which is now the main script used in printing Tibetan works. It further developed into many variants more suited to hand-writing for instance. We can deduce from it's description of the vowel 'e' (which is an inverted triangle representing the yoni) that the Hevajra Tantra still looked to the Siddhaṃ script as its model of writing.

The earliest appearance of Devanāgarī (देवनागरी literally "City of God") is about the 8th century but it did not supplant Siddhaṃ as the main script for writing Sanskrit in North India until about the 10th or 11th century. Many late Buddhist texts, such as later Tantras would have been written in Devanāgarī. Sadly the fluidity of the relationship is lost on most people and the Devanāgarī script is often referred to by the uninitiated as "Sanskrit" - as in "can you right this in Sanskrit for me". Devanāgarī has proved to be remarkably stable - with only minor changes occurring since it was adopted. It has been adapted for writing Hindi and as Hindi is the most prominent of the official languages of India, it is widely used through the sub-continent. Pakistan has adapted the Arabic script fro writing Urdu though it is very similar to Hindi.


Two other scripts which are frequently used in Buddhist contexts are the closely related Lantsa and Ranjana, from Tibet and Nepal respectively. These are often assumed to be identical but this is not true. They emerged in about the 11th century and are both are still in use for ceremonial or decorative purposes. Tibetan texts will often have the Sanskrit title in Lantsa as well as dbu-can at their head.

Although these scripts are all related and all descend from Brahmī originally, knowing one does not always afford insights into reading the others. Some letters such as ṭa stayed remarkably stable, whereas others such as ja changed quite radically over time. Tibetan Uchen retains an archaic form of na which disappeared from India, while other letters are similar to more modern forms, and some appear only loosely based on an Indian precursor. Since conjunct consonants (such as jña or ṣṭha) are written combined into a single glyph and each script does this in slightly different, and sometimes idiosyncratic ways (for instance in Devanāgarī श + री = श्री).

Writing was probably introduced to India by Persian invaders who had themselves absorbed the techniques from other conquered peoples, especially the Aramaeans who were a semitic race that has since disappeared but left a huge legacy. It is thought that parts of the Old Testament of the Bible were composed in Aramaic and that it may have been Jesus's mother tongue. The Aramaic roots of Indian writing are most clear in Kharoṣṭhī. However it was not long before Indians adapted the art of writing to their own languages, and subsequently Indian Buddhists helped to establish the written word in vast swathes of the East including Tibet, Burma, Sri Lanka, and most of South East Asia. Indian scripts influenced the development of the Kana scripts in Japan and the Hangul script of Korea.

Images from Visible Mantra scripts page.