There was a theory that emerged in the early twentieth century which argued that at the end of the first millennium BC, Vietnamese migrated to the Red River delta from an original homeland in what is today southeastern China.

This idea was suggested first by Edouard Chavannes in 1901, and was then developed further by Leonard Aurousseau in 1923. Today this theory is no longer upheld, although one can still find it mentioned. Nonetheless, I don’t think that many people are aware of why this theory does not make sense.

In postcolonial Vietnam this theory came to be rejected as “colonial,” I guess because of its claim that the Vietnamese come from a place outside of Vietnam. However, I don’t find that saying that scholarship is “colonial” explains much.

Yes, there were problems with the scholarship of Chavannes and Aurousseau, but I don’t attribute those problems to a “colonial” outlook. Instead, the problems were due at times to careless scholarship, but also to what I would call a flawed methodology that we can perhaps call the “hidden network approach” to viewing the past.


In a footnote in the fourth volume of his 5-volume translation of Sima Qian’s Historical Records (Shiji 史記), Edouard Chavannes put forth the theory that the “Annamite race” is descended from the people of the ancient kingdom of Yue (Việt). That kingdom occupied the area of what is today the northern part of Zhejiang Province, and it was destroyed in the fourth century BC by the kingdom of Chu.

After that, according to Chavannes, there were multiple polities that formed from “the debris” of the kingdom of Yue, such as Nanyue in the area of what is now Guangdong Province, Minyue in present-day Fujian Province, and Yuedonghai in what is now Zhejiang Province.

vol 4 page 418

Chavannes argues that “these principalities certainly relate to the Annamite race,” and he offers two pieces of evidence to support this claim.

First, he cites Trương Vĩnh Ký’s 1875 Cours d’histoire annamite to claim that Annamite historians regarded the “princes of Nanyue/Nam Việt” as forming the third dynasty of “Annam.”

Second he notes that the capital of the kingdom of Yuedonghai was Dong’ou/Đông Âu (literally, “Eastern Ou/Âu”), and then he cites Gustave Dumoutier’s Étude historique et archéologique sur Co-Loa, capitale de l’ancien royaume de Âu Lạc (255-207 av. J.-C.) (Paris, E. Leroux, 1893), to note that Annamite historians say that there was an Annamite kingdom called Xi Ou/Tây Âu (literally, “Western Ou/Âu) which had as its capital, Cổ Loa, in the Red River delta.

Based on the above information, Chavannes concludes that Dong’ou/Đông Âu, Nanyue/Nam Việt and Xi Ou/Tây Âu were all part of a single Annamite race.


There are numerous problems with Chavannes’s ideas here. First, the “princes of Nanyue/Nam Việt” that Chavanne refers to were Zhao Tuo and his descendents, men who were at least originally what we would call ethnic Han today, that is, people from areas to the north (Zhao Tuo was from Hebei), and not members of the indigenous population of the region, or related to peoples from the ancient kingdom of Yue. So the fact that this kingdom became part of the Vietnamese historical tradition does not mean that it was “racially” the same,

Second, in his 1893 study of the ancient city of Cổ Loa, Gustave Dumoutier stated that “the geography of Gu Xifeng says that under the Zhou dynasty the country occupied by Giao Chỉ was called Lạc Việt, and under the following Qin Dynasty it took the name of Tây Âu or Âu Lạc.” (pg. 8)

Gu Xifeng is Gu Yewang 顧野王, the author of a sixth-century work called the Territorial Treatise (Yudi zhi 輿地志), and the passage that Dumoutier refers to is mentioned in an eighth century annotation to Sima Qian’s Historical Records. This is what it says:

“The Territorial Treatise states that ‘Jiaozhi/Giao Chỉ during the Zhou was Luoyue/Lạc Việt. During the Qin it was Xi’ou/Tây Âu. They tattoo their bodies and cut their hair to avoid serpents. Xi Ouyue/Tây Âu Lạc was to the southwest of Panyu. Yue/Việt and Oulue/Âu Lạc are all surnamed ‘Mi/Mị.’”



This passage is not very clear in that it contains numerous names. However, we can see that Dumoutier “simplified” it by saying that the area of what is now the Red River delta during the time of the Qin Dynasty was called Tây Âu or Âu Lạc.

Dumoutier did not actually say that Cổ Loa was the capital of Xi’ou/Tây Âu. And he did not say that Annamite historians said that either. However, that is what Chavannes said in his book (by citing page 8 of Dumoutier’s book).

So there is definitely some careless scholarship here. Beyond that, however, the ideas that Chavannes expressed are also problematic. First, he equates geographic names (or the names of kingdoms) with race. If, however, we are going to use the idea of race to look at the past, we know that Zhao Tuo was of a different “race” (I think Chavannes was using the term “race” to mean something closer to what we would today call an “ethnic group”) than many of the peoples he ruled over in his kingdom of Nanyue, so it doesn’t really make sense to equate geographic names with single races.

In addition, there is only one source that we can use to make the claim that the Red River delta during the time of the Qin Dynasty was referred to as Xi’ou/Tây Âu (Gu Yewang’s Territorial Treatise), but 1) that text was compiled centuries later, and 2) given that this text contains so many different names, it’s difficult to see that the term “Xi’ou/Tây Âu” is really significant.

However, for Chavannes, the term “Xi’ou/Tây Âu” (Western Ou/Âu) was very important, as it served as a counterpart to another term Dong’ou/Đông Âu (Eastern Ou/Âu).

nodes and links

This then brings us to what I call the “hidden network approach” to studying the past. If we look at the way that Chavannes presented his ideas, it is as if he was discovering a hidden network of nodes and links in the historical record.

The polities – Yue, Nanyue/Nam Việt, and Xi’ou/Tây Âu – as well as the city, Dong’ou/Đông Âu, all served as nodes which Chavannes then discovered the links connecting them.

Dong’ou/Đông Âu and Xi’ou/Tây Âu were linked by the common term “ou/âu,” as well as by the pairing of “east” and “west.”

Xi’ou/Tây Âu and Nanyue/Nam Việt were linked by the fact that Vietnamese historians saw Nanyue/Nam Việt as the third dynasty in their history.

And then since Nanyue/Nam Việt was connected to Xi’ou/Tây Âu and Xi’ou/Tây Âu was connected to Dong’ou/Đông Âu, then it seemed obvious to Chavannes that Yue must be part of this link as well.


I’ve found that scholars like Henri Maspero examined the past in a similar way. To me this approach is overly complex, and it falls apart once people start pointing out problems with some of the links in the network.

So this theory that Chavannes put forth in 1901 had at least three problems with it:

1) It contains some careless mistakes,

2) It is problematic in equating race with place names,

3) It employs an approach to examining the past that is not effective for interpreting the past.

While this theory that Chavannes put forth had these limitations, as I mentioned above, in 1923 Leonard Aurousseau developed it further. In the process, however, Aurousseau did not correct these problems in Chavannes’s theory. I’ll try to write about Aurousseau’s work sometime soon.