Episode 2Era First Chinese domination
Giao Chi, Cuu Chan, Nhat Nam
The Han divided the land of Nam Viet into nine commanderies, yet the table for the year two lists only seven. Of those seven, Giao Chi alone counted more mouths than the other six put together — and to understand that figure, the first thing to ask is what it was counting.
The old chronicles say
What the histories of the dynastic era recorded.
The last episode closed with the setting up of commanderies. The Han thu puts it briefly: having destroyed Nam Viet, the Han made the land into nine commanderies, and it names all nine, from Dam Nhi and Chu Nhai down to Giao Chi, Cuu Chan and Nhat Nam.1
But turn to the treatise on geography in that same book, where its author Ban Gu lists the commanderies of the Viet lands, and you can count only seven.2
Where did the other two go? Looking for the answer, we will meet the three commanderies at the bottom of that list of nine, and a table of households and mouths that can still be read today.
Nine commanderies, then seven
The old chronicles say
What the histories of the dynastic era recorded.
The seven in the treatise are Thuong Ngo, Uat Lam, Hop Pho, Giao Chi, Cuu Chan, Nam Hai and Nhat Nam. Ban Gu writes that all seven belong to the celestial field of the Viet — the stretch of earth that answers to the lunar lodges of the Ox and the Girl.2
Missing are Dam Nhi and Chu Nhai. The same treatise explains: going south into the sea from Hop Pho and Tu Van, one comes to a great island, a thousand li on each side, and the Han made it into those two commanderies. Then the book says plainly why they did not hold: from the time the commanderies and districts were first set up, the Chinese officials and soldiers mostly preyed on and oppressed the people, and so every few years they rose in revolt. Under Emperor Yuan they were given up.2
There is a small discrepancy worth recording. The chapter on the destruction of Nam Viet has the nine commanderies set up all at once, in one hundred and eleven BC.1 The treatise on geography, though, dates the two island commanderies to the first year of Yuanfeng, that is, one hundred and ten BC.2 Two chapters of the same book, exactly one year apart.
The abandonment of the island is told in more detail in the life of Gia Quyen Chi. In forty-six BC this official sent up a memorial arguing against sending troops to Chu Nhai. It is the voice of a man who wants to give the land up: he heaps up every contemptuous word he can find for the Lac Viet, and concludes that they were never worth making into commanderies and districts. The island is damp with mist, full of poisonous plants, snakes and centipedes; before the men have even seen the enemy, the soldiers are already dying of themselves. His closing line is a reckoning of profit and loss: to abandon it is nothing to regret, and not to fight is no loss of majesty.3
Emperor Yuan agreed. His edict placed the reason squarely at home: the lands east of the passes are in great distress, and the granaries and treasuries are empty; sending troops would not only wear out the people, but a year of failed harvests would follow. Then the order: let Chu Nhai commandery be abolished. Those of its people who wished to come over to the Han were to be let stay; those who did not were not to be forced.3
Nine, seven, or eight
The old chronicles say
What the histories of the dynastic era recorded.
So the short answer is: nine commanderies at the founding, seven from the abandonment of the island onward.2 But the histories refuse to be that tidy.
In the Book of the Later Han, the general Ma Yuan himself says at one point that the Han of old sent Lo Bac Duc to open up and set up seven commanderies.4 The Tang commentator Li Xian, reaching that line, felt he had to add a note: here it now says seven commanderies, which does not agree with the earlier book — the earlier book being the Han thu.5
And in Gia Quyen Chi's own memorial, the Han thu has the Han bringing Nam Hai under control and making it eight commanderies.3 The figure eight fits neither of the two dates above, and this episode has not found an explanation for it.
Nine, seven, eight: three figures for the same land, and all three in the official histories. Wherever you meet the words "nine commanderies" or "seven commanderies", the first question to ask is: counted in which year?
The three furthest south
The old chronicles say
What the histories of the dynastic era recorded.
Of the seven commanderies of the Viet lands, the three that lay furthest south were Giao Chi, Cuu Chan and Nhat Nam.
The treatise on commanderies attached to the Book of the Later Han — a work by Sima Biao of the Jin, not by Fan Ye — measures the road down from the capital at Luoyang. Giao Chi lies eleven thousand li to the south. Cuu Chan, eleven thousand five hundred and eighty li. Nhat Nam, thirteen thousand four hundred. The other four commanderies of the Viet lands are all nearer the capital than any of the three.6
Beside the name of Giao Chi, that text also carries a short note that is worth a great deal: this was the country of An Duong Vuong. But the layer it belongs to must be stated clearly. Whether the line is Sima Biao's, or that of the Liang-dynasty commentator Liu Zhao, cannot yet be settled; whoever wrote it, it is a note from somewhere between the third century and the sixth, not the words of the Han thu.6 What it tells us is that people of that time regarded Giao Chi commandery as the old land of An Duong Vuong.
The Han thu lists the districts of each commandery. Giao Chi had ten, among them familiar names such as Me Linh, Chu Dien, Long Bien and Luy Lau.2 Of Luy Lau in particular, a lost geography, the Giao Chau ngoai vuc ky, quoted in the Thuy kinh chu, records that the district was originally the seat of Giao Chi commandery.7 Cuu Chan had seven districts and Nhat Nam five; among Nhat Nam's five is Tuong Lam, a name we will meet again at the end of a sea road.2
Where each district lies on today's map is something the old books help with very little. The Tang commentary, when it needs to point to a place, points to Tang place names: Cu Phong district in Cuu Chan, Li Xian says, is the Ai prefecture of his own day.8 Every mapping onto today's provinces and districts is the work of later scholars.
Ninety-two thousand households
The old chronicles say
What the histories of the dynastic era recorded.
What makes the treatise on geography precious is that it does not just give names. For each commandery Ban Gu records two numbers: households and mouths — that is, homes and people. The whole table is dated to the second year of Yuanshi, under Emperor Ping, that is, the year two AD, more than a hundred years after the commanderies were set up.2
Here are the lines for the three southern commanderies.
Giao Chi: ninety-two thousand, four hundred and forty households; seven hundred and forty-six thousand, two hundred and thirty-seven mouths.
Cuu Chan: thirty-five thousand, seven hundred and forty-three households; one hundred and sixty-six thousand and thirteen mouths.
Nhat Nam: fifteen thousand, four hundred and sixty households; sixty-nine thousand, four hundred and eighty-five mouths.2
The Han thu does not add them up for us. Adding the three lines gives one hundred and forty-three thousand, six hundred and forty-three households, and nine hundred and eighty-one thousand, seven hundred and thirty-five mouths — close to a million people.2
The second sum is stranger still. Giao Chi's mouths alone outnumber those of the other six Viet commanderies put together: the six together come to only six hundred and twenty-six thousand and fifty-three. Nam Hai, the commandery that held Phien Ngung, the old capital of Nam Viet, had only ninety-four thousand, two hundred and fifty-three mouths — not even a seventh of Giao Chi.2
Set against the whole empire, the picture changes. At the end of the treatise Ban Gu sums up the Han realm under Emperor Ping: one hundred and three commanderies and kingdoms, more than twelve million two hundred thousand households, and nearly fifty-nine million six hundred thousand mouths. Then he adds one line: this was the height of the Han. The three southern commanderies make up less than two per cent of those mouths.2
What the numbers counted
How later ages remembered
What later ages did with the story — documents, ceremonies, and arguments of their own time.
These are figures from a court's registers. The natural question: did the registers count real people, or only those who owed tax?
Scholars have taken this up. In nineteen seventy-nine Stephen O'Harrow, in an article setting the Han records beside the archaeology, drew mainly on Hans Bielenstein's analysis of the Chinese censuses. By that analysis the average household across the Han realm held four point seven people. O'Harrow argued that this very ordinary figure is itself the sign that we are dealing with a real census and not just tax rolls.9
Then comes the anomaly. Divide Giao Chi's mouths by its households and you get about eight people to a household. O'Harrow, following Bielenstein, gives the figure for the early Viet population he studied as closer to eight, the highest in the empire; only a few other populous non-Han regions, such as Yunnan or Korea, come close to it. He offers a hypothesis to explain it: the people here were almost entirely non-Han, and may have lived with more people under one roof — he points to the houses pictured on Dong Son bronzes. That is an explanation, not something the sources record.9
Do the same division for the other two and Cuu Chan and Nhat Nam come out at only about four and a half people to a household, no different from the general level.2 The oddity lies in Giao Chi.
Bielenstein's population-density map for the year two, as O'Harrow reports it, gives another picture too: the Red River delta was the most densely peopled part of the Han empire south of the Yangtze, and only Sichuan and the North China plain had a density to match. O'Harrow's conclusion is that the census of the year two offers an independent piece of information, strengthening both the archaeological results and the other texts.9
The limits must be stated as well. Here we are reading Bielenstein only through O'Harrow, and the argument dates from nineteen seventy-nine. How far scholars today trust this table would take newer work that this episode has not been able to read.9
A hundred and forty years on, Giao Chi's numbers are gone
The old chronicles say
What the histories of the dynastic era recorded.
There is a second table, also in Sima Biao's treatise on commanderies, recording the situation under Emperor Shun of the Eastern Han, around the year one hundred and forty.6
There Cuu Chan has forty-six thousand, five hundred and thirteen households, and two hundred and nine thousand, eight hundred and ninety-four mouths. Nhat Nam has eighteen thousand, two hundred and sixty-three households, and one hundred thousand, six hundred and seventy-six mouths. Both are larger than the figures for the year two.6
But for Giao Chi, the text as handed down records only twelve cities, and goes straight on to the list of districts. Where the households and mouths should be, there is a blank; the same is true of Uat Lam.6 That means anyone who gives a population figure for Giao Chi under the Eastern Han did not take it from this table, and should be asked where it came from.
The sea road that starts at Nhat Nam
The old chronicles say
What the histories of the dynastic era recorded.
Nhat Nam had the fewest people of the three. But its name stands at the head of a strange passage in the treatise on geography.
The book says: from the frontier posts of Nhat Nam, Tu Van and Hop Pho, a voyage of about five months reaches the country of Do Nguyen. Four more months by ship reach the country of Ap Lo Mot. More than twenty days' sailing beyond that is the country of Tham Ly. More than ten days on foot reach the country of Phu Cam Do Lo. From there, more than two months more by ship at last reach the country of Hoang Chi. The road is measured not in li, but in months.2
Who travelled that road? The book records a chief interpreter, attached to the Yellow Gate, who went out to sea with men who answered the call, to buy bright pearls, bich luu ly, rare stones and strange things, carrying gold and silks of every kind.2
The book even records one voyage it can put a name and a time to. Around the Yuanshi era, when Wang Mang was regent and wished to make a show of his majesty and virtue, he sent rich gifts to the king of Hoang Chi, and that king sent envoys to present one live rhinoceros. Then the book describes the way back: from Hoang Chi about eight months by ship to Bi Tong, and eight months more to reach the borders of Nhat Nam and Tuong Lam.2
Where Hoang Chi or Bi Tong lie today, the Han thu does not help; it gives only the months at sea. But the point of departure and the point of return both bear the name of Nhat Nam.2
And what does the ground say?
Archaeology finds
Physical traces dug out of the ground.
The books give district names and figures. Does the ground give anything to match them? From the sources this episode has read, the answer is still thin.
O'Harrow, writing at the end of the nineteen-seventies, names the most obvious kind of trace we ought to find if the Han were present in any numbers: underground brick tombs with an earth mound raised over them. Such tombs are found in China itself under the Western Han, and in Vietnam in the later periods of northern rule. But for the time when Nam Viet still ruled, that is, before one hundred and eleven BC, he records that none had yet come to light; the goods in native graves were mostly made locally, from local materials, in local styles.10
That was the state of the evidence as of when he wrote. Much more has been dug since, so the sentence should not be read as describing the situation today.10
One piece of news is more recent. At the end of twenty twenty-four, an English-language bulletin of the Vietnam News Agency reported an excavation at an ancient citadel in Bac Ninh — the bulletin calls it the ancient citadel of Luy Lau — that recovered more than two thousand three hundred fragments of moulds for casting bronze drums, among them an outer mould for a drum face with its decoration complete. But that is a news report, not an archaeological report, and it gives no date for the mould fragments.11 Linking a citadel still being dug to the Luy Lau district that the old books call the seat of Giao Chi commandery is a job for an archaeological report, not for a name in a news bulletin.
So for now, the firmest footing for the three southern commanderies is still a few lines of Ban Gu: three names, twenty-two districts, six numbers, and a sea road measured in months.2
Next episode: Tich Quang, Nham Dien, To Dinh.
Sources for this episode
Every passage points to one of the documents below. How we choose and check sources
Han thu
Used for: The old chronicles sayBack to the text
Han thu
Used for: The old chronicles sayUsed for: How later ages rememberedUsed for: Archaeology findsBack to the text
Han thu
Used for: The old chronicles sayBack to the text
Book of the Later Han
Used for: The old chronicles sayBack to the text
Book of the Later Han
Used for: The old chronicles sayBack to the text
Book of the Later Han
Used for: The old chronicles sayBack to the text
Thuy kinh chu
Used for: The old chronicles sayBack to the text
Book of the Later Han
Used for: The old chronicles sayBack to the text
O'Harrow 1979 — From Co-loa to the Trung Sisters' Revolt
Used for: How later ages rememberedBack to the text
O'Harrow 1979 — From Co-loa to the Trung Sisters' Revolt
Used for: Archaeology findsBack to the text
Viet Nam News, 24 December 2024 — over 2,300 bronze drum mould fragments at Luy Lau citadel
Used for: Archaeology findsBack to the text