Sunday, 24 February 2013

Southeast Asian Scripts and Sanskrit

Everyone on the net seems to be talking about Napoleon Chagnon.  I have nothing to add to this debate and it has descended into name-calling.  A lot of the same accusations are flying again - Chagnon hates the Yanomami, Neel and Chagnon had Yanomami people kill one another in exchange for axes and money, etc. - despite their having been debunked repeatedly.  It's boring and annoying, so I'm going to focus on something else: southeast Asia inscriptions and Sanskrit.

The earliest inscriptions in southeast Asia are in Sanskrit.  Actually, that's only sort of true - the earliest inscriptions are definitely in Sanskrit, but they were followed so closely by native languages (specifically, Cham) that to emphasise the Sanskrit epigraphy would be misguided.  Nonetheless, from Myanmar through Thailand, Cambodia, Vietnam, and out to Java and Bali, Sanskrit appears to have served as a liturgical language, if not something more, for at least a few hundred years.  Sanskrit and, eventually, Pali, have had such clear impacts on modern southeast Asian languages that they must also have carried some weight as vernaculars as well, although clearly not in the same way as Latin did.

If you look at Malay, you'll see a huge Sanskrit influence, especially in the Indonesian dialects of it (including standard Indonesian).  Commonly-used words like bisa (meaning 'to be able to, can') from Sanskrit bhisa ('power') inflect daily spoken language in southeast Asia with an Indic superstrate.  So Sanskrit wasn't just as a religious language; it had, and continues to have, a broader impact (Sanskrit is still used to form new words in Indonesian today, just as Latin and Greek are in English and other European languages).  Tamil is also represented in inscriptions from the same period, and had a similar impact on southeast Asian languages, which is not surprising given the significant interest of Tamil traders in the region, and given the fact that Indic culture appears to have been introduced to Indonesia and beyond by south Indians.

Georges Coedes, an early twentieth century scholar of southeast Asia, noted that Sanskrit was first written down in southeast Asia almost as early as it was in south Asia itself.  There was very little time lag between Sanskrit's written use in India and its use in epigraphy in Borneo and Vietnam.  It also seems likely that there was a lag of less than a century between Gautama Buddha teaching in Magadha (c.4th century BCE) and the introduction of Buddhism to southeast Asia (indicated by the presence of lion figures carved from Indian carnelian in Thailand, specifically Ba Don Tha Phet, at a time when the lion represented Gautama himself).  At this early stage, parts of south and southeast Asia appear to have formed a continuum of overlapping cultural groups united at the upper echelons by the spread of Indian religious traditions from centres in north India, and perhaps by the use of Sanskrit.

More recently, the Indologist Sheldon Pollock, of Columbia University, has claimed that Sanskrit used to serve as the 'cosmopolitan language' of the area from Kashmir to Vietnam and Indonesia among populations who spoke diverse languages.  He sees Sanskrit as a unifying influence within this sphere prior to the arrival of Timur in India in the fifteenth century.  I find this troubling for a number of reasons, the foremost of which is that while at the beginning of the period (early mid-first millennium CE, roughly c.300-400) the script used in all of southeast Asia was the same (the Pallava script, written with variations both on the mainland and in the maritime regions), after the first century or two there was a sudden diversification and regionalisation in scripts.  It is also clear that the original script for Sanskrit in southeast Asia came from south India, and was actually very different from scripts used in north India and Afghanistan in the same period.  This means that at no point in the history of Sanskrit would a text written in Borneo or Cambodia have been immediately legible to somebody in Kashmir, and vice versa.

At the same time, it is clear that there was a diversity of scripts in use in the islands, at least, in the late first millennium (Kawi, a script used to write Old Javanese and Old Malay, is clearly not a direct development of Pallava, and obviously has several antecedents).  There are north Indian inscriptions in Nagari from the Melaka Straits dating to c.900-1000 CE, indicating a bit of mixing.  Still, this use of diverse scripts and intense regionalisation and so-called 'indigenisation' within a few short centuries indicates that Sanskrit was likely an oral cosmopolitan language if it was a cosmopolitan language at all, and given that the probable transmission of Buddhist thought to Thailand took place at such an early date - before, indeed, any Indic language had even been written down in India (discounting the evidence of written Old Indic in the Mitanni empire in Bronze Age Syria) - it seems as if Sanskrit didn't need writing to be transmitted at all.  In this it differs markedly from Classic Mayan, Latin, and Classical Chinese in spreading primarily through oral means.

Either way, Sanskrit does appear to have had a brief life as a religious and political language in the region.  It died out in southeast Asia in response to a number of significant changes, not least of which was the general abandonment of Hinduism.  On the mainland, Theravada Buddhism's uptake by local elites led to the growth of Pali, Thervada's liturgical language, as the higher tongue, a position it has, to some extent, maintained to the present.  In Indonesia, Sanskrit continued in inscriptions until about 1500 CE, when the Majapahit state came to an end in the wake of the introduction and spread of Islam.  It is also clear that the Majapahit form of Hinduism was competing with, or was mixed with, various Buddhist traditions, including Vajrayana ('tantric' Buddhism, now confined primarily to Tibetan Buddhism), and with local Javanese beliefs that had presumably originated with the founding Austronesian-speaking population of the island.

If Sanskrit indeed served as the cosmopolitan language of southeast Asia at any point, it didn't last long in this role, and throughout its existence it was threatened by local languages (Indic 'prakrits', Dravidian, Austroasiatic, Austronesian, and Tai-Kadai, etc.) and competing religious and social traditions with Indic origins. It is also notable that the first written language in India wasn't Sanskrit - it was Magadhi Prakrit, the language of the Mauryan empire, a Sanskrit-derived Indic language.

Sanskrit's spread was uneven, and most texts in southeast Asia derive from a single Indic antecedent, where India itself has many recensions and variations.  See, for instance, the Ramayana; in India, there are lots of versions of the Ramayana, and there appear to have been lots of versions for as long as the story has been in existence.  In Java, however, there have also been lots of versions, but they have all consisted of local embellishments of a single Sanskrit text, the Bhattikavya, a version of the Ramayana composed in verse form designed as a teaching aid for Sanskrit grammar.  It is notable that this composition is memorable enough to serve as an oral text, and need not have been transmitted purely through Sanskrit texts of Indian origin.  It does not appear that Sanskrit was used as the principal means of transmission for texts across the entirety of its sphere, and, as I said, a text from Kashmir would have required transcription in order to be read in Java.

I tend to look at southeast Asia through indigenous goggles.  My interests are primarily prehistoric.  But I think this provides a better means of looking at the region than the 'Indianisation' view, which sees all southeast Asian states and literary forms as little more than epiphenomena of Indic influence.  It is pretty clear to me that Indic influence was neither unified nor sufficient to overrule the great diversity of local traditions, and I doubt the theory that Sanskrit was a truly cosmopolitan language in the way Greek once was in the eastern Mediterranean.  I don't believe that traders could have sailed from Gujarat to Vietnam via Melaka without changing language along the way, and I question the view that Sanskrit was the unambiguous 'cosmopolitan' language of so-called 'Greater India', whether in the mid-first millennium CE or at any point before or since.


  1. Thank you for this post. I speak and read Cambodian at a functional level, and have little to no familiarity at all with Sanskrit or Pali, save through the Cambodian dictionaries I use that note Khmer words of Pali or Sanskrit origin. This is an interesting take on them.

    I do not know if this is directly relevant to the written vs. oral transmission model, but from my (admittedly subjective) viewpoint, there seems to be more Khmer words of Sanskrit than Pali origin - or at least, I have used more words from the Sanskrit side of things .

  2. I don't think the model is one of either oral or written transmission, but a mix. In the case of Sanskrit, there is an associated oral literature (the Vedas, and probably the Upanisads, Astadhyayi, Dharmasutras, etc) and doesn't seem to have spread through the use of any particular script, which makes me think that it was principally orally transmitted. Most languages are orally transmitted anyway - but it's odd that a prestige language should spread like this. Akkadian didn't become the diplomatic language of the ancient Near East by being transmitted orally to Egyptian scribes, and Latin wasn't the prestige language of Europe for a millennium because it was being spoken in the street, or even because of being spoken at all (hence the great divergence in Latin pronunciation in different European languages).

    As for Pali vs Sanskrit, it seems to vary a lot. Indonesian has almost no Pali words, because Theravada was not an important branch in the archipelago (it has a lot more Arabic and Persian, in addition to Sanskrit). Thai has lots of Pali, because Tai-Kadai speakers arrived in mainland southeast Asia in large numbers as late as the twelfth century CE, after the peak of Sanskrit influence. I would have thought Khmer would exhibit more variation, but it's interesting to know that it has more Sanskrit than Pali.

  3. By the way, I wrote a post a while back about the Austroasiatic language family, of which Khmer is a part:

  4. "I would have thought Khmer would exhibit more variation, but it's interesting to know that it has more Sanskrit than Pali."

    Please never quote me on that! I have no data to back it up. There are plenty of Pali words in the dictionary - but I rarely use them. On the other hand, I use Sanskrit words enough recognize many of the Sanskrit roots. I don't know any Pali roots. I suspect that this is because Pali words are reserved for high Buddhist thought or perhaps the 'monk level' of language, which I am not very familiar with. Much more Sanskrit seems to have made its way down to every day speech.


You can post anonymously if you really want to, but I'd appreciate it if you could provide some means of identifying who you are, if only for the purpose of knowing who has written what.