As I stated in the previous post (The “One Country – One King” Problem in Premodern Southeast Asian History), when historians have looked at the records of tribute missions to China from Southeast Asian polities in the past, they have generally viewed those records as representing tribute missions from individual kings from individual countries.
I, however, do not view the information about tribute missions in that way. Instead, I think that while we do have evidence at times of tribute coming from one king in one country, at other times we can see that foreign merchants (such as Cham merchants) delivered tribute for a Southeast Asian king, and at still other times we can see that foreign merchants “usurped” the tribute relationship and delivered “fake tribute” from a certain polity for their own benefit.
This is very clear in the records about tribute missions during the Song Dynasty period (960-1276), and particularly with the tribute records about “Kambuja.”
In 1983, historian Robert M. Hartwell compiled a list of information about tribute missions during the Song Dynasty period. In this document, Hartwell assumed that records about a place called “Sanfoqi” were about “Srivijaya,” that records about “Zhenla” were about “Cambodia,” and that records about “Shepo” were about “Java.”
By this point, I have determined beyond a shadow of a doubt that this is not accurate. Instead, “Sanfoqi” referred to “Kambuja,” “Zhenla” referred to an area in the Lower Mekong region (or in the Mekong Delta), and “Shepo” referred to the area around what is now Songkhla on the Malay Peninsula.
While I think that I have provided ample evidence of this already, seeing the information about tribute missions presented in this list format of Hartwell’s document makes it even clearer.
Historians have long believed that “Zhenla” was the name that the Chinese used for “Cambodia,” and during the Song period that meant “Angkor.” Angkor at that time was one of the biggest human settlements in the world, and yet. . . according to Hartwell, there were only 2 tribute missions from Zhenla, and neither of them mention a king. . .
That makes absolutely no sense.
Zhenla was not Angkor. Sanfoqi was. So, let’s see what the Song-era records say about Sanfoqi.
The earliest Sanfoqi/Kambuja king mentioned in Song Dynasty sources is someone whom Hartwell felt the Chinese characters for his name indicated “Udayadityavarman.” This is the same name as that of a person whom, as we saw in the previous post, George Cœdès argued was a Cambodian king for a few months around 1001.
As I stated in the previous post, Cœdès had no evidence to support this idea. All we can really determine from the available inscriptions (as far as I can tell) is that a ruler named Udayadityavarman existed.
What we see in Song dynasty sources, meanwhile, is the existence of a king named Udayadityavarman from as early as 960 AD.
Then in the 980s, we find records of a Sanfoqi/Kambuja king who has a title that seems to include the term “Haji.” Does this mean that Udayadityavarman had been overthrown by a foreign “invader” or that some Khmer ruler had converted to Islam and had come to power?
I doubt it. What I think is more likely is that it is evidence of foreign traders interfering in the tributary relationship between Angkor and China.
You can get more of a sense of that with a record from 985 which refers to a “Master of Ships” from Sanfoqi/Kambuja. What is this “master of ships”? My guess is that it was a reference to a position like a “shahbandar,” a person who was in charge of trade and collecting taxes on trade.
When I see these records, I imagine an international entrepôt where certain ethnic groups have their own communities, and which are led by a member of their own community whom the local king has appointed.
The sudden appearance of a “Haji” and a “Master of Ships” in the 980s after a period when tribute arrived on behalf of a king Udayadityavarman suggests to me that something was happening in Kambuja. It looks like foreign traders were taking control of the tribute missions. Is that because there was political turmoil and Udayadityavarman had lost control (or had less control)?
A few years later, we see a clearer sign that things were not going well for the political elite in Angkor. In a record from 988, we learn that Sanfoqi/Kambuja had been invaded by Shepo (which Hartwell has as “Java” based on a mistaken idea that Shepo and the later Java were the same place), an important trading polity on the Malay Peninsula that had a long-running rivalry with Sanfoqi/Angkor.
Then in 1003, there is a tribute mission from Sanfoqi/Kambuja on behalf of a king called Sri Chulamanivarmadeva and in 1008 from a king Sri Maravijayottungavarman. These two people are mentioned in inscriptions from the Chola kingdom in southern India. I will write more about this later, but the gist is that I would argue that these are the names of kings who controlled the trans-peninsular trade route from Kedah to Songkhla.
That area was all under the authority of Shepo, so these people were likely the kings of Shepo. It looks like during this time period these two men tried to extend their control over other areas. One inscription from the Chola kingdom indicates that Sri Maravijayottungavarman assumed control of Kedah. If his father (or someone before him) attacked Angkor, then this would indicate that this ruling family was trying to bring the entire trade network from Kedah to Angkor directly under their control.
This attempt to control this trade network did not last because the Chola kingdom attacked the areas around the Malay Peninsula that the Shepo king controlled.
Why did the Chola kingdom do that? It’s not clear, but we have evidence in other Chola inscriptions of gifts from “Kamboja,” so perhaps someone from Angkor convinced the king of Chola to stop the effort of the Shepo kings to dominate trade in the Southeast Asian region.
Did the Chola kingdom get anything in return beyond these gifts? It would seem so, and subsequent information from the eleventh century gives us a clue as to what happened.
After tribute arrived in 1017 from a (claimed) ruler with the title of “Haji,” in 1028 and 1077 the title “Deva” appears.
Hartwell was confused by the entry for 1077 and argued that it must contain erroneous information as in that year a Chola king is recorded to have presented tribute, and someone with the same name was appointed by the Song court as a “Great Chieftan” of Sanfoqi/Srivijaya.
I do not think that this is a mistake. Instead, I would argue that this might point to the Chola kingdom’s domination of Angkor’s trade in this time period. Or perhaps more likely, what this could also point to is the presence of Tamils from the Chola kingdom who stayed in the region after the Chola attack on Shepo and who essentially became the main “shahbandar” for Angkor. As they did so, they may have both pretended to deliver tribute for the Chola kingdom as well as carry out tribute missions for Angkor.
Two years after the Song appoint a “Great Chieftan” of Sanfoqi/Srivijaya, a mission arrived from the “Sanfoqi Zhanbi” kingdom. Hartwell, following the idea that Sanfoqi referred to a place called “Srivijaya” on the island of Sumatra, rendered “Zhanbi” as “Jambi,” a place on that island.
However, the History of the Song makes it clear that “zhanbi” was not a place name. Instead, it states that “zhanbi” 詹卑 is what the king of Sanfoqi was called.
In the paper I wrote on this topic, I argue that this is a rendering of the Khmer word for a ruler, “somdech” (a term which we find in inscriptions from that time period), as the term “zhanbi” would have been pronounced 1,000 years ago as something closer to “tchambec” (I’m writing this from memory, but anyone can look it up in Pulleyblank’s reference work).
Does it make sense that in 1079 a mission would have been sent from the “Kambuja Somdech’s Kingdom”? DEFINITELY!!! Because we can see before this point that someone who was not the king had been granted a special position, and I would argue that this person was not Khmer, and was perhaps Tamil.
In other words, someone was trying to make a clear distinction by saying “We represent the real king of Kambuja!!”
In the years that immediately followed, we can see that this “Deva” person, whoever he was, continued to deliver tribute, so there could have been relations between the Song Dynasty and both the king of Kambuja and the shahbandar as well.
What I have just presented is what I would call a “first look” through these sources. It’s important to consult the primary sources, and it is also extremely important to examine and talk about the envoys who delivered tribute.
Nonetheless, what I think the above material clearly shows is that we cannot look at tribute from a place like Sanfoqi/Kambuja and imagine that it represents the tribute from “one king” in “one country.”
It was much more dynamic and complex than that. Angkor was an international entrepôt, and in many periods foreign merchants appear to have dominated Angkor’s economic relations with the outside world.
At the same time, when we realize that “Sanfoqi” was “Kambuja” rather than an imagined “Srivijaya,” we can resolve issues that scholars have never been able to resolve, such as issues surrounding “the Chola invasion of Srivijaya”. . .
However, before we address that topic, let us look at the information in Arabic sources that is relevant for this topic.
