Srivijaya 2.0 (2): Where was Sanfoqi?

Sanfoqi is the Chinese name of an historical kingdom that scholars believe once existed on the island of Sumatra and was called “Srivijaya.”

I disagree. I am 100,000% convinced that Sanfoqi referred to “Kambuja,” that is, the Cambodian empire based at Angkor.

One of the clearest ways to see this is to look at where Chinese authors geographically placed Sanfoqi.

Sanfoqi was first recorded in Chinese sources during the Song Dynasty period (960-1279). Its location was indicated by noting its position vis à vis other polities.

So, for instance, there was a kingdom mentioned in Song-era sources called Nanpi 南毗, and its location was indicated in reference to Sanfoqi. Zhao Rukuo’s thirteenth-century Treatise on the Various Barbarians (Zhu fan zhi諸蕃志) stated of Nanpi that: “It is in the far southwest. With favorable winds, it can be arrived at from Sanfoqi in just over a month” (Zai xinan zhi ji; zi Sanfoqi bianfeng, yueyu ke dao 在西南之極;自三佛齊便風,月餘可到).

This comment indicates that Nanpi was over a month away from Sanfoqi, and the sense is that it was to the southwest. If we follow the standard interpretation that Sanfoqi = Srivijaya = Palembang (Sumatra), then the only place a month away to the southwest from there is the Kerguelen Islands, also known as the Desolation Islands, in the southern part of the Indian Ocean. . .

That is probably not where Nanpi was. Instead, the “far southwest” in Song times could have indicated the island of Sumatra. In which case, if Sanfoqi was over a month to the northeast of there, that would put it right around. . . Cambodia.

The History of the Song, which was actually compiled during the following Yuan Period (1279-1368), states of Sanfoqi that “Neighboring Champa, it is between Zhenla and Shepo” (Yu Zhancheng wei lin; ju Zhenla, Shepo zhi jian 與占城為隣,居真臘、闍婆之間).

“Neighboring Champa” is easy to figure out, and that would place Sanfoqi again in the area of Cambodia. As for being between Zhenla and Shepo, let’s try to figure out what that means.

“Zhenla” is a Chinese name that scholars think refers to “Cambodia.” I am convinced that it referred to the lower Mekong region, basically somewhere between Phnom Penh and the coast where the Mekong enters the sea.

For our purposes here, let’s just accept that Zhenla refers to someplace in Cambodia. If a polity is “neighboring Champa” and is between Zhenla and another place, then one conclusion we can make is that it can’t be all that far away. In other words, this can’t be referring to Palembang on the island of Sumatra, as the island of Sumatra was not “neighboring Champa.”

Then there is this other place, Shepo. Where was that? In the previous series on Srivijaya, I argue that this was a reference to a polity on the Malay Peninsula in the area of what is now Phatthalung and Songkhla provinces in Thailand, and particularly the area of Sathing Phra.

I still think that is the case, but even if we ignore that point, and just base our understanding on the idea that Sanfoqi was “neighboring Champa” and between Zhenla and another place, that information alone still forces us to locate Sanfoqi on the Southeast Asian mainland.

Let us now look at a source from the Ming Dynasty period (1368-1644). In the fifteenth century, some Ming scholar-officials compiled a geographical text known as the Gazetteer of the Unified Realm of the Great Ming (Da Ming yitongzhi 大明一統志).

In that work, Sanfoqi was said to be “a 5-day journey to the south of the kingdom of Champa” (Zai Zhancheng guo nan wu ri cheng. . .在占城國南五日程).

Here again, we are definitely not talking about the island of Sumatra, as we can find references in this same text that indicate that it would take at least 20 days to reach Sumatra from Champa.

Finally, I came across a map from 1618 that was created based on information in the Gazetteer of the Unified Realm of the Great Ming. This map is not “modern” in that it doesn’t locate places precisely in geographic space. Instead, it places kingdoms in relation to each other.

As for Sanfoqi, this map places it right next to Zhenla and An Nam (i.e., Vietnam). Here again, this is definitely not indicating the island of Sumatra.

Therefore, from its first appearance in Chinese texts during the Song period to the end of the Ming period, Sanfoqi was repeatedly located right near Champa, and far, far away from the island of Sumatra.

This is clear evidence that Sanfoqi was not Srivijaya.

Share This Post

Leave a comment

This Post Has 2 Comments

  1. Historia Mundi

    But then again, Zhu Fan Zhi clearly states that the people of Sanfoqi make alcoholic beverage out of flowers, coconuts, and areca which is a very austronesian thing (kamboja is not austronesian). Zhu Fan Zhi also states that the people “live on the water on rafts of boards covered over with reeds” this is a traditional houses of palembang people up until 1900s when the dutch still rule over indonesia.

Leave a Reply