It all began with Sanskrit inscriptions. . .

Over the past few posts, I have been examining (and deconstructing) information in Chinese sources that has long been used to support the idea that there was an historical kingdom called Srivijaya that was based on the island of Sumatra and that had “maritime supremacy” over much of Southeast Asia from roughly the seventh to fourteenth centuries AD.

The idea that there once was such a kingdom was first proposed in 1918 by French scholar George Coedès. He said that there were a couple of names in Chinese sources that referred to this polity, Shilifoshi and Sanfoqi, and in the previous posts I demonstrated that those terms do not refer to a place that we can think of as Srivijaya. While I still have some questions about Shilifoshi, there is no doubt in my mind but that Sanfoqi was Angkor.

What I would like to do in this post, is to look at where Coedès got the idea that there was a kingdom named “Srivjaya” in the first place. He got that idea from reading some Sanskrit inscriptions.

I studied Sanskrit for three years, and I was a bad student, so I only have a rudimentary knowledge of that language. Nonetheless, as limited as my knowledge of Sanskrit is, I think I can demonstrate that we can find problems with Coedès’s claim that there was a place called Srivijaya.

On the one hand, we can admire and sympathize with people like Coedès and others who were the first to examine inscriptions, such as Georges Maspero, Étienne Aymonier, and Abel Bergaigne. It’s not easy to read a faded inscription, and there were limited resources for learning Sanskrit and Khmer and Cham a century ago, so in that sense what those scholars accomplished in producing transcriptions and translations of inscriptions was definitely impressive.

On the other hand, however, it is also clear that these men read inscriptions with a “mental model” in their minds. They read inscriptions looking for political information about the historical relations between kingdoms. In particular, they wanted to know the names of kings and kingdoms, and they wanted to put together a chronology of what had happened in the past.

The problem is that many of the extant inscriptions are not about “politics.” They’re religious in nature, and even those that do contain political information tend to refer to rulers in divine ways.

Before Coedès wrote anything about Srivijaya, a Dutch epigrapher by the name of Johan Hendrik Caspar Kern wrote about an inscription that was found in 1892 in Kota Kapur on the island of Bangka off the coast of Sumatra.

That inscription is in Old Malay and it contains the expression “kadatuan srivijaya.” “Kadatuan” is a term that we can translate as “polity,” and this expression can therefore be read to mean the “polity of Srivijaya.”

But who or what was “Srivijaya”? Kern understood Srivijaya to mean “His Majesty Vijaya.” In that sense, “kadatuan Srivijaya” would have meant “the polity of His Majesty Vijaya.”

In 1918, however, Georges Coedès disagreed. He argued that Srivijaya was the name of a kingdom.

Coedès, like his colleagues at that time, was looking for kings and kingdoms, and in his mind, a term like “srivijaya” had to indicate EITHER a person (Sri Vijaya, or “His Majesty Vijaya”) OR a place (Srivijaya, “Le Royaume de Srivijaya”).

This was a very modern/Western way of reading ancient, and essentially divine, texts. Coedès was looking for the equivalent of “the Louis I of Sumatra” and “the France of Borneo.”

But what I will try to argue below is that nothing like this is in the inscriptions in Southeast Asia. Instead, they are filled with divine titles. As such, as I see it, “Srivijaya” is NEITHER the name of a person NOR of a place. It is a status, that of “the Victorious Lord” or “the Great Victorious One.”

From that perspective, “Kadatuan Srivijaya,” can easily be “the polity of the Victorious Lord.”

So Coedès was looking for kings, but his argument about Srivijaya was based on Sanskrit grammar. In Sanskrit, there are different types of compound words. In his article, Coedès mentions two types: karmadharaya and tatparusa compounds.

The rules for compounds are way more complex than I can explain here, but I’m going to put them in the simplest possible terms here for the sake of clarity.

In karmadharaya compounds, the first word functions kind of like an adjective and modifies the second word (the lion king), whereas in a tatparusa compound, the second word is more important and “rules over”  the first (the king of lions).

Coedès used this distinction to argue that Srivijaya had to be understood as the name of a kingdom rather than the name of a king.

He made his point by discussing a Sanskrit inscription that had been found in the area of what is today southern Thailand. That inscription dated from 775 AD and in it were three titles that contained the word “Srivijaya,” but Coedès made his argument based on only the following two:

Çrïvijayendrarâja (Srivijaya + Indra + raja)

Çrïvijayeçvarabhûpati (Srivijaya + Ishvara + bupati)

So we have “Srivijaya,” then we have either the god Indra or Ishvara (a term that can mean “god” in a more general sense or that can be a reference to Vishnu), and then the Sanskrit words for “king” (raja) and “lord” (bhupati), respectively.

According to Coedès “In Indochinese epigraphy, Indra and Ishvara are never used as the second term in a karmadharaya compoud with the sense of ‘king-’: for example, one never encounters a form like Jayavarmendra(râja) or Jayavarmeçvara(bhûpatï) to indicate ‘the king Jayavarman.’”

(Dans l’épigraphie indochinoise, indra ou ïçvara ne sont jamais employés comme second terme d’un composé karmadhâraya, avec le sens de « roi-» : par exemple, on ne rencontrera jamais une forme telle que Jayavarmendra(râja) ou Jayavarmeçvara(bhûpatï) pour désigner « le roi Jayavarman ».)

Therefore, he argued, since the title “Çrïvijayendrarâja” has “Indra” as the second term in a compound, and with the sense of “king,” then this title has to be read as a tatparusa compound and translated as “king (of the country) of Srivijaya” (roi [du pays] de Çrïvijaya).

He argued further that this “country of Srivijaya” was the same as the Srivjaya mentioned in the Kota Kapur inscription that Kern had written about.

And with that Sanskrit grammar argument about the way to read two titles in an inscription from southern Thailand that dates from 775 AD, “the maritime empire of Srivijaya” was born. . .

Ok, so the first point I would make is that the Kota Kapur inscription that Kern wrote about, as well as the various other inscriptions that have been found since that time in southern Sumatra that contain the term “Srivijaya” are all in Old Malay.

As such, “Srivijaya” does not appear in those words as part of a Sanskrit compound that we can apply karmadharaya and tatparusa rules too.

Therefore, Coedès’s argument that there was a kingdom called Srivijaya is based on a Sanskrit grammar argument that he made about two titles in a 775 AD inscription from southern Thailand. That’s not a very strong argument.

Let’s just imagine though that he is correct. The inscriptions from Sumatra that contain the term “Srivijaya” all come from basically the same time period: The Kota Kapur inscription (686), The Kedukan Bukit inscription (683), the Telaga Batu inscription (683), and the Talang Tuo inscription (7th century).

If “Srivijaya” in these inscriptions is, as Coedès argued, indicating a kingdom, then it is very odd that the inhabitants of that kingdom only had the urge at one point in history to document their kingdom’s existence.

If, however, the “Srivijaya” in those inscriptions was referring to a person, as Kern originally argued, then it would make much more sense that a powerful individual, someone who took the title for himself as “the Victorious Lord,” would want to document his power during his lifetime.

Let’s return though to Coedès’s argument and particularly the comment that “one never encounters a form like Jayavarmendra(râja). . . to indicate ‘the king Jayavarman.’”

(on ne rencontrera jamais une forme telle que Jayavarmendra(râja). . . pour désigner « le roi Jayavarman ».)

What is Coedès translating here – Jayavarmendra(râja – as “the king” (le roi)?

He says that “Indra and Ishvara are never used as the second term in a karmadharaya compound with the sense of ‘king-’.” Ok, so if “endra” (Indra) in this example means “king” then what about “raja”? That also means king. Why does Coedès put that term in parentheses?

What I think we see here is Coedès’s modern/Western fixation on (or obsession with) finding “the king” (le roi). He is seeing two things in this title, a king and something else. But there are more than two things in these titles.

To get back to Sanskrit grammar, what we have here is not a two-word tatparusa compound, as Coedès translates it (le roi Jayavarman). Instead, it’s a multi-word compound.

What are the rules for multi-word compounds?. . . Ahhhh, I told you I was a bad student of Sanskrit!! I can’t really remember, but I think the way it works is that words form into compounds and then combine with other words or compounds.

So with, Çrïvijayendrarâja, the title in the 775 AD inscription, you could have (Srivijaya + Indra) + (raja) or you perhaps could have (sri + vijaya) + (Indra + raja). I’m not sure if “sri” was considered a separate word when it comes to compounds as it is so commonly connected to nouns (and that’s the type of thing that Sanskrit grammarians love to argue endlessly about. . .).

So how do we translate Çrïvijayendrarâja? I would translate it as the “Great Victorious Indra King.” Such a rendering does not enable us to see who “the king” actually was, but we can find countless titles like this in “the Indochinese inscriptions.”

I think I have probably lost a lot of readers by this point (if they are still reading). However, that is an important point.

The idea that there was a maritime empire called Srivijaya is based on an argument about Sanskrit grammar that is written in French.

How many people have tried to figure out what exactly Coedès was saying? My guess would be that it’s not very many, and of those that did, there were probably none who had the ability to then go on and check if the Chinese terms that Coedès said referred to Srivijaya (Shilifoshi and Sanfoqi) actually do.

The result has been that in the century since Coedès made his claim, people have just taken his word for it.

That’s what I did. I entered the world of scholarship learning that “George Coedès discovered Srivijaya.” I assumed that must be true, so I never felt the need to read his 1918 article, especially since it’s in French (Ouch!).

So it’s only a week or so ago that I first read this article. What I think this post hopefully makes obvious is that it takes a lot of work to try to figure out Coedès’s argument. However, once one does, it’s shocking to see how weak it is.

Indeed, he simply doesn’t have an argument at all.

So yes, we now know from some Old Malay inscriptions that in the 680s there was a datu in southern Sumatra who was referred to as “the Victorious Lord” (Srivijaya), and it certainly looks like he was one tough dude.

But as for Coedès’s maritime empire of Srivijaya, there is no evidence of that, and Coedès himself never provided enough evidence for anyone to take his claim seriously either.

He just got lucky that the people who read his article either didn’t know Sanskrit grammar or couldn’t read texts in classical Chinese or both.

Now after a century his luck has run out.