By now I think that I have clearly established the fact that there was a kingdom located in the area of what is now Songkhla that was at the center of an important trans-peninsular trade network.
I refer to this kingdom as “Jāba” (although it may have actually been called “Jāva”) and have found evidence of it in multiple sources, where it is referred to variously as Jāba/Zabaj/Shepo (pronounced something like “Jāba” in the past). Yet, one more place that we can find reference to it is in the main Vietnamese chronicle the Complete Book of the Historical Records of Đai Việt (大越史記全書 Đại Việt sử ký toàn thư), compiled in the fifteenth century.
I am going to present those references here. However, to do so requires that we first understand a rather complex linguistic issue. So, here we go!!
The earliest mention of this place that I call “Jāba” (in order to differentiate it from the island of Java) comes from the History of the [Liu] Song (comp. 492-3) which records that in 435, a king from a place variously called Javasad(h)a 闍婆娑達 or Javavad(h)a 闍婆婆達國 sent tribute to the Chinese court. The seventh-century History of the Southern Dynasties then mentions Javad(h)a 闍婆達, whereas later sources then all just mention the first two characters, 闍婆.
These early names that we find in Chinese sources were coined by Chinese monks who made the journey to India. They were familiar with the way that Chinese characters were used to transliterate Sanskrit terms, and they at times applied that approach to transliterate placenames in Southeast Asia, some of which, like this one here, appear to have been Sanskrit-based.
That said, we cannot always reproduce with certainty the intended transliteration, and in the case of 闍婆, this could have been Java or Jaba, or even Yava or Yaba, as the first character, 闍, was used to represent both “ja” and “ya” in Sanskrit, and the second character was used to represent both “va” and “ba.” So, it’s possible that the original four-character name could have been Jabasad(h)a or Jabavad(h)a or Jababad(h)a, etc. It’s anyone’s guess.
So, these characters were used to represent Sanskrit terms, but they also had “normal” pronunciations in Chinese that were slightly different from the sounds in Sanskrit they were used to transliterate. What is more, those pronunciations changed over time as the language changed.
As we move later in time, when only the two characters, 闍婆, were used, there were probably fewer people who read these two characters with knowledge of the technique of transliterating Sanskrit terms into Chinese. These people would have read these characters based on how they were pronounced at that time, and there were actually two different ways to pronounce the first character.
Today, this character, 闍, is pronounced as either “she” or “du.” Those two pronunciations are the end result of linguistic transformations over time. According to Edwin G. Pulleyblank’s reconstruction of historical pronunciations in the Tang, Song and Yuan periods, the two pronunciations of this character would have changed as follows (see the image below for the phonetic symbols):
The current “she” pronunciation = dzia / shia / sɛ
The current “du” pronunciation = tɔ / tuə / tu
Meanwhile, the second character, 婆, which is now pronounced as “po,” would have changed as follows: ba / phua / pɔ
Therefore, in the Tang, Song, and Yuan periods, this is how 闍婆 may have been pronounced by someone who read those two characters:
Tang: Dzia-ba or Tɔ-ba
Song: Shia-phua or Tuə-phua
Yuan: Sɛ-pɔ or Tu-pɔ
Today: Shepo or Dupo
Having said all of that, in the case of the Chinese, it doesn’t really matter how a person thought those characters were pronounced because it was an established way of referring to a certain part of the world and people just kept using those characters.
Once the name of a tributary kingdom was established, that same name was used in Chinese sources for as long as tribute from that part of the world kept arriving at the Chinese court, regardless of whatever political and ethnic changes may have taken place on the ground.
As such, a term like “Jiaozhi” 交趾 for the area of what is now the northern part of Vietnam was used for a good 1500 years, during which time there were plenty of transformations on the ground.
Therefore, the fact that the name, Shepo/Dupo 闍婆, had two different pronunciations, and pronunciations that changed over time, didn’t really matter to the Chinese. If tribute came from that place, they used that name. That’s all that mattered.
For Vietnamese, however, the story is different, and this is what we need to understand before we examine information in Vietnamese historical sources.
Prior to the tenth century and the emergence of an autonomous “Vietnamese” polity in the Red River plain, there was no “Vietnamese” historical tradition (as far as we can tell). It is only after that point that we have evidence of people in the area of what is now Vietnam creating dynastic chronicles to record information about the past.
Later, in the fifteenth century, the Complete Book was compiled and it included information from before the tenth century. That information, however, came largely from Chinese historical sources and was put in a separate section, known as the “Outer Annals,” in contrast to the sections about Vietnamese dynasties and based on information recorded by Vietnamese themselves, known as the “Main Annals.”
Why is this important? It is important because in contrast to the Chinese, who had a set term that they could use, Shepo/Dupo 闍婆, whenever there was contact with people from the Malay Peninsula around Songkhla, the early Vietnamese kingdoms did not necessarily have such a term.
If the person who recorded information about contact with such people in the dynastic chronicles was aware of the earlier Chinese records about Shepo/Dupo 闍婆, then he could have used that term, but if he was not aware of that information, then he would have had to come up with a name on his own.
You may be completely confused by this point, but I’ll try to explain below why we need to think of this when we read the Vietnamese sources that contact information about Shepo/Dupo 闍婆, or what I call “Jāba.”
Here is the earliest mention of Jāba in the Complete Book.
767 – “Côn Luân/Kunlun 崑崙 and Chà Bà/Shepo 闍婆 raided. Military Commissioner Zhang Boyi requested assistance from Commander-in-chief of Vũ Định, Gao Zhengping. After the assisting troops arrived, the Côn Luân/Kunlun and Chà Bà/Shepo/Jāba forces were defeated at Chu Diên. Boyi rebuilt La/Luo Citadel.” (Ngoại kỷ [Outer Annals], 5/4b-5a)
丁未唐代宗豫大曆二年。崑崙闍婆來寇，攻陷州城。經畧使張伯儀求援於武定都尉髙正平。援兵至，破崑崙闍婆軍於朱鳶。伯儀更築羅城。
Đinh Mùi [767], (Đường Đại Tông Dự, Đại Lích thứ 2). [Người] Côn Lôn, Chà Bà đến cướp,
đánh lấy châu thành. Kinh lược sứ Trương Bá Nghi cầu cứu với Đô úy châu Vũ Định là Cao Chính Bình.
Quân cứu viện đến, đánh tan quân Côn Lôn, [5a] Chà Bà ở Chu Diên. Bá Nghi đắp lại La Thành.
This term, “Chà Bà/Shepo 闍婆,” is the same term that is used in Chinese sources to refer to the place near Songkhla that I refer to as “Jāba” (to differentiate it from Java).
That is not surprising because this account comes from Chinese records, and was only included in the Vietnamese chronicles when the Complete Book was compiled in the fifteenth century. Therefore, this information, although it is historically the earliest, actually entered the Vietnamese historical chronicles later than the records that we will look at next.
Did the person who included this information look at the characters, 闍婆, and pronounce them as “Chà Bà,” as the person who translated this passage did? Probably not.
Whoever translated this passage was aware that Vietnamese in more recent times have referred to the Malays in the Mekong Delta as “Chà Và,” and this is how that name was Romanized starting in the nineteenth century.
However, the Vietnamese pronunciation of the character, 闍, is either “xà” or “đồ.” Therefore, the scholar who included this information in the Complete Book in the fifteenth century most likely looked at those two characters and pronounced them something like “Xà Bà” or “Đồ Bà.”
As for the other terms here, “Côn Luân/Kunlun” is a reference to Côn Đảo/Pulo Condore, a group of islands to the southeast of the southern tip of the Indochinese Peninsula. In Tang-era sources this place was referred to as Mount Juntunong (軍突弄山), and in Arabic sources, it was referred to as Kundrang or Kadrang.
Meanwhile, “Chu Diên” was the name of an area in the Red River plain, and La/Luo Citadel is the name of a citadel in the area of what is now Hanoi, a precursor of the Thăng Long Citadel built in the eleventh century.
In Chinese records, the area that we refer to as “Champa” went by different names: first Linyi, then Huanwang, and finally Zhancheng. I bring this up to point out that there wasn’t a “stable” term in Chinese sources for referring to people in this part of the world at the time this record was made, and that therefore, reference here to Côn Luân/Kunlun is most likely a reference to some group of Cham(ic) peoples, and does not necessarily mean that they were literally from the islands of Côn Đảo/Pulo Condore.
However, from what I have found, it looks like there was a close relationship between Jāba and Cham(ic) peoples. This reference to the two attacking the Red River plain supports that idea.
1066 (9th lunar month) – “A merchant from Trảo Oa 爪哇 presented a nocturnally-radiating pearl, and the price paid for it was 10,000 coins.” (Bản kỷ [Main Annals], 3/4a)
爪哇商人獻夜光珠，還價錢萬鏹。
Lái buôn người nước Trảo Oa dâng ngọc châu dạ quang, trả tiền giá 1 vạn quan.
1149 (2nd lunar month) – “Merchants from the three kingdoms of Trảo Oa 爪哇, Lộ Lạc 路貉, Xiêm La 暹羅 entered the Eastern Sea and requested to take up residence to trade. A settlement was established on islands in the sea for engaging in trade and presenting goods [to the court] called Vân Đồn 雲屯.” (Bản kỷ [Main Annals], 4/6b)
己巳十年宋紹興十九年春二月，爪哇、路貉、暹羅三國商舶入海東，乞居住販賣，乃於海島等處立庄，名雲屯，買賣寳貨，上進方物。
Kỷ Tỵ, [Đại Định] năm thứ 10 [1149], (Tống Thiệu Hưng năm thứ 19). Mùa xuân, tháng 2, thuyền buôn ba nước Trảo Oa, Lộ Lạc, Xiêm La vào hải Đông6, xin cư trú buôn bán, bèn cho lập trang ở nơi hải đảo, gọi là Vân Đồn, để mua bán hàng hoá quý, dâng tiến sản vật địa phương.
Vân Đồn is the area where the famous Hạ Long Bay in northern Vietnam is located. It was a site where foreign merchants congregated, and here we have merchants coming in 1066 and 1149 from Trảo Oa 爪哇.
These two characters are the exact characters that Chinese eventually came to use, starting in the late 13th century, to refer to the island of Java. If you look at early Ming sources, like the Veritable Records of the Ming (Ming shilu 明實錄), there is another character that sometimes gets used for the second half of this name, but eventually, the term gets stabilized to using these two characters (Zhuawa 爪哇).
So, is Trảo Oa 爪哇 here a reference to the island of Java? Or is it how Vietnamese recorded the name of the place at Songkhla and it’s just a coincidence that Chinese would later use the same two characters when they started to record information about the island of Java in the late thirteenth century?
Theoretically, it is possible that people from the island of Java went all the way up the eastern side of the Malay Peninsula and then across to Champa and then up the coast to Vân Đồn. However, given that,
1) there are historical connections between the Songkhla “Jaba/Java” and Champa, and given that
2) the other places mentioned in the record from 1149 are from the Gulf of Thailand region (see below), and given that,
3) I can’t find evidence of other foreigners referring to the island of Java as “Java” until the late 1200s and early 1300s, and given that,
4) the characters used here are standard characters for representing the sounds of foreign words (more on this below), and given that,
5) the earliest reference from the island of Java that I have been able to find of people from there referring to that place as “Java” is the mid-fourteenth-century text, the Nagarakretagama, and given that,
6) I can see that the main international trade route in this period is still going over the Malay Peninsula at places like Songkhla rather than through the Straits of Melaka,
I strongly suspect that this is a reference to the Songkhla Jāba, and that Vietnamese created this term without knowing that there was already an established term in Chinese sources.
These records come from the early years of the Vietnamese historical tradition and were recorded around the time that these events actually took place. These were records that the Vietnamese created themselves. If the person who recorded this information didn’t know that there already was a term in Chinese sources for these people, then he would have had to create his own term, and he would have done so in a way that came close to representing the sound of the name.
As for those other places, Xiêm La 暹羅 is “Siam,” and Lộ Lạc 路貉 seems to have been a place on the eastern side of the Malay Peninsula not all that far to the south of Bangkok today, and perhaps around where Chumphon Province is.
As we saw in the post in Marco Polo, he visited a place around that area which he referred to as “Lochac.” The second character in the name in the Complete Book, 貉, is usually transcribed as “lạc” in Vietnamese (that’s a long story), however, it should be pronounced as “hạc,” and “Lộ Hạc” gets us very close to Marco Polo’s “Lochoc.”
So, again, theoretically it is possible that merchants traveled all the way from Java to northern Vietnam. And if they did, and this term referred to them, then this would be the earliest historical evidence for the use of the name Java to refer to the island of Java.
However, I think it is much more likely that this Trảo Oa is a reference to the other “Java” in the Southeast Asian past, what I call the Jāba at Songkhla.
1184 (3rd lunar month) – “Champa came to present tribute. Merchants from the kingdoms of Xiêm La and Tam Phật Tề/Sanfoqi entered Vân Đồn, presented precious goods and requested to engage in trade. (Bản kỷ [Main Annals], 4/20a)
甲辰九年宋淳熙十一年春三月，占城來貢。暹羅、三佛齊等國商人入雲屯鎮，進寳物，乞行買賣。
Giáp Thìn, [Trinh Phù] năm thứ 9 [1184], (Tống Thuần Hy năm thứ 11). Mùa xuân, tháng 3,
nước Chiêm Thành sang cống. Người buôn các nước Xiêm La và Tam Phật Tề vào trấn Vân Đồn dâng vật báu để xin buôn bán.
There is no mention of Jāba here, but we do get the only mention in the Complete Book of a place called “Tam Phật Tề.” This is “Sanfoqi,” the name that George Cœdès claimed was a reference to a place called “Srivijaya” on the island of Sumatra, and a term which I have demonstrated is a Chinese name for “Kambuja,” that is, Angkor.
As I discussed in a previous post, it looks like Angkor’s trade was largely carried out by foreigners. Therefore, it is not entirely surprising to see this account from 1184 implying that this was the first time that someone from Angkor had tried to engage in trade at Vân Đồn, a place that had been established as a trading site a few decades earlier.
It could be that foreign traders from Angkor had arrived earlier, but that this was the first mission to identify itself as representing Angkor.
On a different note, the fact that traders had to first present precious goods to presumably government authorities in order to gain approval to engage in trade is an indication of one of the ways in which places that served as trade hubs became wealthy.
This helps us see how wealthy the trade hubs on the Malay Peninsula could have been, and why so many different peoples wanted to control them.
1348 (10th lunar month) – “A merchant ship from the kingdom of Đồ Bồ 闍蒲 arrived at the settlement of Vân Dồn to secretly purchase oyster pearls [there is a note here about how this term is pronounced]. Many people stole floating (?) pearls and gave them [to the merchants from Đồ Bồ. When this was discovered, all were punished.” (Bản kỷ [Main Annals], 7/14a-b)
冬十月，闍蒲國商舶至雲屯海庄潜買蠙(蠙音駢，蛛也，蛛又音朱，亦作黿，)珠，雲屯人多偷氽珠與之。事覺，俱扺罪。
Mùa đông, tháng 10, thuyền buôn nước Đồ Bồ đến hải trang Vân Đồn ngầm mua ngọc trai. Người Vân Đồn [14b] nhiều kẻ mò trộm ngọc trai bán cho họ. Chuyện này bị phát giác, đều bị tội cả.
1349 (5th lunar month) – “The kingdom of Đại Oa [‘Great Oa’] came to present local goods as tribute, as well as a talking red parrot.”
夏五月，大哇國來貢方物，及能言赤鸚鵡。 (Bản kỷ [Basic Annals], 7/14b)
Mùa hạ, tháng 5, nước Đại Oa sang cống sản vật địa phương và chim vẹt đỏ biết nói.
This term, Đồ Bồ 闍蒲, seems to again be a reference to the place at Songkhla that I call Jāba. The first character is the same character that is in “Shepo” 闍婆, the old Chinese name for that place. However, the second character here, 蒲, is different and is pronounced: “bồ.”
The translator of this passage has transliterated the first character as “đồ,” but as we’ve seen, that character is also pronounced as “xà.” So, the person who made this record may have tried to write something like “Xà Bồ.”
What I find interesting here is the use of the first character 闍. When it comes to producing the sound “ja,” this is not an obvious choice to use to represent that sound. The records from the eleventh and twelfth centuries use the characters “trảo“ 爪 and “oa“ 哇. These are standard characters for representing sounds like “ja” and “va.”
闍 is not, particularly given the fact that it has two pronunciations. The reason why this character was used to represent a “ja” sound in the case of “Jāba” is that this was done around the fifth century following a system that was used to represent Sanskrit with Chinese characters. In later centuries, people did not commonly use this character to represent a sound in a foreign name.
Therefore, the person who recorded this information, must have been familiar with the old term “Shepo” 闍婆, but for some reason did not reproduce the same second character (maybe because what he heard or reported sounded differently?).
Finally, as for the second record above, I would argue that “Đại Oa” (“Great” Oa) is definitely a reference to the island of Java, and more specifically, to the kingdom of Majapahit. By 1349, Majapahit had already extended its influence far beyond the island of Java. Further, I argue in the second half of my paper (coming out early next year), that Majapahit attacked Angkor in the 1370s and occupied areas of the lower Mekong region for several decades after that point.
That Majapahit would send a mission to Vietnam to make its presence known at this point strikes me as extremely logical.
These are the references to “Jāba” and related places in the Complete Book. In 1349, Vietnamese definitely encountered people from the island of Java. Prior to that point, while theoretically, it may have been possible for such an encounter to take place, my guess (especially given all of the other information I have found) would be that they only knew the Jāba/Jāva at Songkhla, and that they recorded that name in different ways over time.
This Post Has 4 Comments
A fascinating post as usual! Also, this makes me wonder about a few things:
1. Do I understand correctly that Javanese did call their island “Java”, at least from the period of and as evidenced in Nagarakretagama? I ask this because I did read somewhere that “Java” was a total outsiders’ name for the island.
2. Also about Java, what is your assessment about the “Iabadiu” or “Jabadiu” from Ptolemy’s Geographia? I have just read about it on the Wikipedia page for Java and I cannot find any further reading about this (of course, my search is very limited due to non-expertise).
3. About the Vietnamese name “Cha Va”, as far as I know, it’s a Romanized name referring to Malay people (Javanese included). However, I wonder how it came to be. Could it be, as you suggested, there had been a name for “Jabanese” used by “Vietnamese” in the Mekong Delta and it had been written in Han or Nom characters by officials, then overtime it was transformed into the Romanized “Cha Va” and broadened in meaning to include all Malay people? FYI, there is an old area in the former Cho Lon named “Cha Va” and the current theory is that Cholonese mistakenly called Indian (or Tamil) merchants having been frequent there as Javanese (from the island). Perhaps two things can be true: Cholonese mistakenly called the Indian as Malay, so they used the old term and the term is now interpreted as referring to Javanese.
Bonus: I’m very curious about the “Lac” story are teased in the post. Is there anyway I can read about it now? Or is another post warranted for that?
Hello there,
1) Yea, I’ve read scholarly articles that say that, however, the name does appear in a few inscriptions. There is an article available online which talks about them: ”
The problem of the ancient name Java and the role of Satyavarman in Southeast Asian international relations around the turn of the Ninth century CE” by Arlo Griffiths.
2) As for Ptolemy. . . and then there are ancient Indian texts as well that mention names like Yavadvipa (Yava/Java island). . . I haven’t looked into those. My guess though would be that such names were created without knowledge of what “locals” thought or called the place they lived (like the way the term “America” was first used), but that later, as the elite in places like the island of Java became literate in Sanskrit, they started to use those terms for themselves.
3) That the popular term in Vietnamese is now “Cha Va” instead of something like “Jaba” could be because of contact with Khmer. In Khmer this same name is more like Cha Va. The historical contacts between Jaba/Songkhla were more with places in the Mekong Delta, and Vietnamese only migrated into that area relatively late. So, yes, it could be that this pronunciation was adopted from Khmer. I think Truong Vinh Ky may have even made that point somewhere.
Bonus: Oh, I think what I meant there is how there is another term which sounds the same, Lac Viet, which the Vietnamese have used to create a sense of the antiquity of the Vietnamese nation. A colleague and I recently wrote an article on that, and it has been translated into Vietnamese too:
https://www.academia.edu/62405577/Competing_Imagined_Ancestries_The_L%E1%BA%A1c_Vi%E1%BB%87t_the_Vietnamese_and_the_Zhuang
https://www.academia.edu/82825564/Tranh_ch%E1%BA%A5p_c%C3%A1c_v%E1%BB%8B_t%E1%BB%95_t%C6%B0%E1%BB%9Fng_t%C6%B0%E1%BB%A3ng_Ng%C6%B0%E1%BB%9Di_L%E1%BA%A1c_Vi%E1%BB%87t_ng%C6%B0%E1%BB%9Di_Vi%E1%BB%87t_v%C3%A0_ng%C6%B0%E1%BB%9Di_Choang
Thank you for the reply. As I read the article and some of the references, I wonder if you believe Vu The Ngoc’s hypothesis that the Chinese characters for “Lạc” were basically used to represent the sound of the word used by “Lac people” (whatever that means) at least since Han dynasty, and that word meant “water”.
Come to java island
We have many inscriptions about khmer mon cham people.
In kaladi inscription on java 906 ce wrote : ariya dravida pandhya kling shinghala khmir mon cham and hunjeman (arab persia rom merchants) are foreigners who must pay tax as workers or traders in java archipelago.
They cant go to city or villages just in the ports.