If you look at old European maps of the area of what is now southern Thailand, you will notice something odd. From the seventeenth to the nineteenth centuries, these maps make it look like there was a large island along the coast.
And then at the end of the nineteenth century, that island suddenly disappears, and we get maps that essentially show what that part of the world looks like today, with its coastal strip of land and the inland lakes of Thale Noi, Thale Luang, and Lake Songkhla.
I’ve been struggling to understand this for a long time, but now I think I have figured it out. Essentially, I think that 1) this is an area that Europeans did not really investigate clearly but they knew that it was possible to travel through the interior by water, and 2) in the absence of new, more accurate information, cartographic information tended to get replicated from one map to the next.
So, someone early on depicted on a map an interior water route from Nakhon Si Thammarat to Songkha, probably based on hearsay rather than direct observation, and that’s how everything stayed until more updated information was finally incorporated in the late nineteenth century.
The reason why I have come to think this way is because there is an early-nineteenth-century Vietnamese text which clearly documents the passage through the interior, and the information it provides demonstrates how inaccurate the depictions in European maps before and after that time were.
That text is called the Collected Records of the Routes in the Kingdom of Siam (Xiêm La quốc lộ trình tạp lục 暹羅國路程集錄), and I’ve written about it here on the blog before.
The final text in this collection is called the Estuary Water Routes (Hải Môn thủy trình 海門水程), and in discussing a few places in what is now southern Thailand, it provides detailed evidence of the inland water route between Nakhon Si Thammarat and the interior lakes: Thale Noi, Thale Luang and Lake Songkhla.
Let us take a look.
六坤海門，門廣約二十尋，水深三尺，兩傍林藪，民居田疇，江長屈曲，水程一更，至六坤茫。此茫有瓦城，暹官鎮守，兵約三千率，民居稠密，所藝亦如(礻/占)奔。自六坤城江長屈曲，兩傍民居林雜，以至窮源。
The Lục Khôn estuary is approximately twenty xun [fathoms] wide, with a water depth of three chi [Chinese feet]. On both banks, there are forests and thickets, as well as dwellings and cultivated fields. The river is long and winding. A water journey of one watch [2.4 hours] leads to Lục Khôn mương.
At this mương, there is a brick-walled citadel garrisoned by Siamese officials, commanding approximately three thousand troops. The population is dense, and their trades are similar to those of Chiêm Bôn. From Lục Khôn Citadel, the river winds for a long distance, with dwellings intermixed with forests on both banks, continuing all the way to the river’s source.
Notes:
1) Lục Khôn 六坤 is “Nakhon,” and refers to Nakhon Si Thammarat.
2) Mương 茫 is the Tai term, “mueang” (เมือง), meaning a polity.
3) Chiêm Bôn (礻/占)奔 is “Chumphon,” and is mentioned earlier in the text.
I think this part is pretty self-explanatory, so let’s move on to the next record.
泊坡能海門，門廣約二十尋，水深二尺，兩傍林藪，江長屈曲，水程一日，江分爲兩派，俗名我𠀧北烈。一派向西，江長屈曲，兩傍民居林雜，以至窮源；一派向南，江長三曲，兩傍民居田疇椰葉以至分守。自分守小江，江長屈曲，兩傍茅葦林雜，水程一日至泑羅喎海門。
The Bạc Pha Năng estuary is approximately twenty xun wide, with a water depth of two chi. On both banks, there are forests and thickets. The river is long and winding. After a water journey of one day, the river divides into two branches, commonly known as Bắc Liệt Fork.
One branch flows westward; the river is long and winding, with dwellings intermixed with forests on both sides, extending all the way to its source. The other branch flows southward; the river winds through three bends, with dwellings, cultivated fields, and coconut palms on both sides, reaching Phân Thủ.
From Phân Thủ, traveling along the small river, which is long and winding with reeds and mixed forests on both sides, it is a one-day water journey to the Ao La Oa estuary.
Notes:
1) Bạc Pha Năng 泊坡能 is “Pak Phanang.”
2) For “Fork,” the text has the Vietnamese term, “ngã ba” 我𠀧, meaning “intersection” or “fork.”
3) Bắc Liệt 北烈 – I’m not sure exactly where this was, but “bắc” should be “paak” (ปาก) in Thai, meaning “mouth,” as in a “river mouth.”
4) Phân Thủ 分守 – I’m also not sure where this was, but the “phân” could be “baan” (บ้าน) in Thai, meaning a village.
5) Ao La Oa 泑羅喎 – The “ao” is “aao” (อ่าว) in Thai, meaning a “bay.” La Oa is “Rawa,” and refers to a place on the coast now known as “Ban Pak Rawa.”
The above image indicates where the above passage started (Pak Phanang) and ended (Ban Pak Rawa).
泑羅喎海門，門廣約四尋，水深一尺，俗名丐飭，兩傍林藪，水程二曲，江分為兩派，一派向西北，通注我𠀧北烈，一派向西南，江長屈曲，兩傍茅葦茂盛，通入渃𠮾𤅶。
The Ao La Oa estuary is approximately four xun wide, with a water depth of one chi. It is commonly known as Cái Sắc. On both sides, there are forests and thickets. The water route bends twice, and the river divides into two branches: One branch flows northwest, connecting and pouring into Bắc Liệt Fork; the other branch flows southwest, where the river is long and winding with lush reeds and rushes on both sides, leading into the Sweet Water Sea.
Notes:
1) Cái Sắc 丐飭 – I have no idea what this is.
2) Sweet Water Sea is a Vietnamese name, but written in Chinese word order, Nước Ngọt biển 渃𠮾𤅶.
This passage takes us in the opposite direction. While the above passage traced the route from Pak Phanang to Ban Pak Rawa, this passage documents the route moving inward from Ban Pak Rawa, and it mentions a place where one could turn to the northwest to reach Bắc Liệt Fork or turn to the southwest to reach the Sweet Water Sea.
My sense is that this is the same place as Phân Thủ, mentioned in the previous passage, from where one took the “small river” to go to Ao La Wa/Ban Pak Rawa.
You can open this Google Earth map to view what is depicted in the above image more closely.
What is very clear when we examine the landscape in Google Earth is that to the north of Ao La Wa/Ban Pak Rawa, a large canal has been constructed, and this has clearly affected the area to its south, as has agricultural expansion.
This is why I have pins in orange and yellow. The orange pins are along places where there still are waterways, while the yellow pins are in places where there previously were waterways. Indeed, if you zoom in on the brownish area on the map, you can clearly see the traces of large rivers that once flowed through that area.
At that intersection that I have labelled as Phân Thủ, it looks like there was a river that went directly south, so, it is possible that this was the route to the lakes.
All of this is difficult to say with certainty as that canal has totally transformed this area.
Nonetheless, what I have documented on that map is very close to what is recorded in this historical account, and this account is by far the most detailed account of travel in this region that I have ever seen.
Finally, the next place mentioned in this text is Songkhla. The text does not indicate that one got there by traveling inland, as of course, it was accessible by sailing down the coast.
However, the interior lakes were all connected, so it was possible to travel all that way by water, and this is what I assume Europeans learned early on and why for centuries their maps depicted water from Nakhon Si Thammarat to Songkhla.
Another thought provoking post.
1) was there an island, or was it an illusion?
The geomorphological and archaeological studies suggest Satingpra peninsula that trapped the sea and created the lagoons (lake songkhla) and the area to the north is a fairly recent creation, originally started as a barrier island that eventually connected pre-quarternary islands like koh yai, kao daeng, some 4000 bp.
The area north of thale noi, once a channel or strait, is also believed to be recent, and remained boggy swamp, as CJ Reynolds describes.
“ The channel between
Tantalem Island and the mainland became shallower and shallower to
the point where oceangoing vessels could no longer navigate through the
passage and the three ‘lakes’ gradually took shape.”
“By the end of the nineteenth century, much of the basin at the northern
end of the lakes was a mosquito-infested swamp forest (pa phru) as the
channel silted up.”
( The mid south’s fathomable past, 2019 p67, 68)
John Walker’s map to go with John Crawfurd’s 1828 book, clearly depicted Tantalem island.
Up to the mid 19th century the strait was wide and good enough for a 1400 ton naval vessel to go through as described by FA Neal (1852). When W Smyth visited the same area fifty years later, the strait was gone (1896). It shows how quickly the strait was silted up.
2) canals
With the ever disappearing strait, some rivers and klong could have “emerged” in the landscape naturally, and the people may learned from them and extended them to preserve transport corridors. There must have been many waterways, to go either side of Mt Prabat(another pre-quarternary rock, once an island), coming from Pak Panang via Cha-uat to Thale noi (as the visitors from Vietnam went through).
The ones going the eastern side was depicted in Warington Smyth’s map as Ranawt river. Curiously the shape of the land cut by the river resembles with Tantalem island depicted in some western maps from 18 and 19th centuries. (Ranawt, or Ranot, Ranod, btw is my consideration for ao la oa, could it be? Ranot being probably the earliest settlement on the island/peninsula, it’s worth considering.)
3) How it was depicted on the western maps.
It seems the island (unnamed) appeared first in the mid 17th century, almost coinciding with the listing of a settlement called Singora. The western sailers must have arrived.
As you say, some maps just copied existing ones. Yet in the 18th century on,
the island got names, Pulo Papier, Tantalam and Tantalem (at least by 1775), Ko yai (a former island, now part of Satingpra peninsula) and one 1835 map, Talung-lem(lem being a headland, could mean Patalung headland?).
The shape of the island vary, and as I mentioned above, some look very close to that of the land separated by river Ranawt in W Smyth’s map.
There are some 19th century maps that depicts Tantalem as a peninsula, with a river to Talung while the other connects the peninsula with the mainland in the south. So there were some curious variations in the depiction of Tantalem, in the western eyes, I think, as the shape of waterways shifted.
Thank you very much for these extremely helpful comments!!!
1) I think someone might have mentioned Reynold’s piece to me before, but I didn’t have access to it at that time. So, thank the dark powers that be that I was now able to find it online. . .
I have been trying to imagine some gradual process of silting over the centuries, but that somehow didn’t make sense to me. Reynolds, however, makes this point:
“Until the late nineteenth century, vessels could pass unimpeded through the bay, which gradually filled as the runoff from
the mountains deposited sediment in the lowlands. Forest land on the plains and slopes was cleared for agriculture and, with the construction of roads and rail lines, silting in the bay accelerated. The channel between Tantalem Island and the mainland became shallower and shallower to the point where oceangoing vessels could no longer navigate through the passage and the three ‘lakes’ gradually took shape.”
Smyth (Five Years in Siam, vol. 2, pg. 90) mentions jungle men burning the forest to clear rai (fields) for planting. I don’t know what scale of forest burning one would need to lead to the kind of silting that seems to changed that region rather quickly, but perhaps that was it, as I have never come across references to major logging operations in this area at that time, like you had further north in the teak forests.
2) I found Smyth’s map here: https://illustrationarchive.cf.ac.uk/page_turner/3427182/000351/2
If you zoom in with a browser you can see it more clearly. It looks like in this map, the southernmost stretch of the route to the lakes is the one that I mentioned above that we can see on Google Earth went directly south from where I’ve put “Phân Thủ.”
However, Smyth’s map was made almost a century after the Vietnamese traveled through this area, so that southern route might not have existed at that time, and could be something that opened up as the southwestern route to Thale Noi silted up.
3) As for La Oa being Smyth’s Ranawt. . . From what I can see, today there is a district called Ranot, with the district seat on the land to the east of Thale Luang. Meanwhile, Rawa is a subdistrict of Ranot District, and it is on the coast.
That “oa” in “La Oa” is definitely transcribing “wa” rather than “not” and it has “ao” before it, meaning “bay” or “cove.” Further, what I have translated as “estuary” is literally “sea gate.”
So, there seems no question to me that this was a reference to the place on the coast that is today identified on Google Earth as “Pak Rawa,” with “pak” meaning “mouth” as in a river mouth.
Meanwhile, on Smyth’s map, he has “Pak Ranawt” on the coast right where “Pak Rawa” in the Ranot District is on Google Earth.
Seeing all of that, I think what we are seeing here are signs of the “Thai-ification” of place names that came with the establishment of the modern Thai nation and its administration.
Ranot comes from Rawa, and is a Thai pronunciation of that placename. It looks like now both exist, but the Thai one is in the place where a modern district administrative center has been established (in perhaps 1924?).
Smyth’s map comes from a period shortly before that time, and is reflecting a Thai pronunciation of Rawa, but before there was a district of that name, with a district seat in a different location, inland from the coast.
Instead, in addition to Pak Ranawt, he also has a “Tung Ranawt” meaning a Ranawt plain or expanse, thung (ทุ่ง), where the filled in area to the north of the lakes is, and a “Klong Ranawt,” or Ranawt canal going through that area. So, Ranawt in his day was different from the Ranot district of today, but it looks like Pak Rawa/Ao La Oa/Pak Ranawt all consistently point to the same place on the coast.
4) You make a brilliant point about how the channel on Smyth’s map matches the shape of Tantalem in some European maps. Yes! That’s a great observation. In some maps it bulges out toward the west in one part. I’ve long wondered why it was depicted that way, but now that makes sense.
Again, thank you so much for these extremely helpful comments!!!
Yep, it took me a while to understand the geomorphology of the area, and when we look at the history of the area, we have to be really careful and imagine what the land looked like in those days. It was definitely not what it is now. Jane Allen’s main criticism of J Stargardt’s work on Satingpra is exactly that.
The barrier island and its high ridges act as a wall to trap sediments from western mountain ranges and accelerate the silting process.
What was it like for example, when it was raided by Chola in the 11th century for example?
Satingpra was still very much an island and its centre (trade and admin) or the citadel was believed to be at chatingpra, until the 14th century. While Stargardt detected some “disturbances” in the 9th century, her archaeological studies did not show any signs of a foreign raid in the 11th, I recall.
Singanakhon the seat of Singora sultanate is believed to be a new town set up in the 17th. What was there before we don’t know, whether there had been a settlement or indeed a fort.
S Wavell quoted a local education officer at patalung as saying”Satingpra was deserted twice. About 300 years ago, people fled to khao Daeng. The second time to patalung.”(p196)
So, something happened in the 16th century that drove the people to south of the island in singanakhon, giving rise to the sultanate in the 17th.
When it was sieged and torched by the invading ayutthaya forces, Singanakon was ruined so much that the French declined to purchase when offered by Bangkok in the 1780s.
The population was believed to have moved to Lem son (Laem son) on the same island. Lem son is noted by some maps, while the Singora at singhanakhon or kao daeng doesn’t seem to have appeared at all.
As I mentioned above it was only early 17th when the western sailers came to know about singora, Segora was first mentioned by robert larkin in 1614 when he sailed in Patani. Whether he actually went there or not we aren’t sure, but it was there and estimated to be as good as batavia.
Meanwhile Mueang Songkhla in Bo Yang, opposite of Lem son, is said to have been built much later, its founding was mid 19th. If it was a fresh settlement built in the 19th, where exactly was Singora before, as shown in the western maps of 18th century?
3) Ranot, Ranod, Rawa
Yes, the thai-fication of the place names is really annoying and complicating things. And the official spelling doesn’t match with the local pronunciation. Locals call it RanoD, though officially it’s spelled with a T, for example.
It could well have been Rawa was the original name as you say, and may have covered the wider area. We find south of Ranot, there’s ban rawa, a beach called rawa, but this ban doesn’t seem to have an immediately obvious port.
This greater Ranod area, once the northern tip of the emerging Tantalem, was believed to have been inhabited (by settlers from where? Funan, or Talung, which was then located further down south at the time?) as early as 3rd century. Khok thong east west canal was in operation at least by the 6th century. At the mouth of this canal we have Utapao, the port, as a gateway.
Utapao is also the name of klong that would take you south towards kedah. Suspect it was also how people called a settlement around Hajai those days, as the name of the first railway station was called utapao junction.
4) Great to know that someone else is staring at the canals and the area, old maps trying to understand the history that at least makes sense. So much is still untold.
Keep your good work.
More very helpful comments. Thank you again!!
Back 1,000 years ago, I think most places of any significance were up a river. It was too dangerous to be right on the coast. Given that the trade route to Kedah was clearly the most direct and most important, a place like what is now Hat Yai seems like the ideal location for a polity.
I just looked into the name Utaphao, and there is a story that the u is อู่ = a dock, cradle, berth, or place where something is kept/built, and the taphao ตะเภา is an old word for a Chinese junk / sailing ship, equivalent samphao to สำเภา
My “invented tradition” radar went off when I read this. My guess would be that this is a later explanation, and that it is a Thai-ification of a name from another language.
The 1225 Zhufan zhi says that Shepo was called Pujialong 莆家龍. The phonetic reconstructions for those characters for 1,000 years ago are not all that different from how they sound today.
It’s a stretch, but I can get the vowels of u-a-ao to coincide with u-ia-o [I have in my head the Vietnamese pronunciation of “long” which is more like “lowng and is closer to an “ao”].
Again, it’s a stretch, but both Pujialong and Utaphao are distortions of foreign words. Could they be distortions from different languages and different centuries of the same word?
One final thought: I keep looking at that area in Google Earth and thinking to myself what an archaeological and treasure hunting goldmine it is. One could start by flying some remote sensing device over those traces of former rivers/canals to look for signs of submerged boats, and then do the same for the lakes. At the very least, there must be ceramics there.