Sungai Batu, Bujang Valley, Kedah, Malaysia
Bujang Valley
Iron smelting in the early centuries CE (an older sample treated as an outlier)
Under a red roof in a Kedah oil-palm grove, iron furnaces are labelled third to fifth century. A 2005 claim put an older lost city in a Johor forest. The expedition that went to see it reported no city.
Why this is extraordinary
Two early Malaysias are in circulation, and they are not the same place. In the Bujang Valley of Kedah, between Gunung Jerai and the Muda River, excavation has opened an iron-smelting industry and a set of brick shrines. A 2020 study of the furnace site called SB2A dates the smelting to the first to fifth centuries CE by radiocarbon, by accelerator mass spectrometry, and by optically stimulated luminescence, the light-clock in grains that have been buried and then dug into the sun. The furnace ran as a bloomery, iron made in the solid state, at about 1150 to 1200 degrees Celsius, with tuyeres, the clay nozzles that let the blast in, and with haematite brought from Bukit Tupah. Slag, ore, brick, beads, and ceramics come with the furnaces. The valley, on that account, was not only a harbour for other people’s goods. It made iron.
A later reading of the radiocarbon set, reported in 2023 from the same university’s archaeologists, says a Bayesian model of the samples puts most of them in the second to tenth centuries CE, in line with the buildings and the written history, and treats one sample of 788 to 537 BC as an outlier unless more samples of that age turn up. The shelter in the photograph is labelled third to fifth century, which sits inside the first study’s span and inside the later model’s span, and does not include the eighth century BC. An outlier is a sample that the model cannot use without breaking the rest. It is still a sample. If more like it appear, the model moves. If they do not, 788 BC is a date without a city.
The brick shrines are the other half of the valley. The second photograph is a restored candi base, stepped brick in a park clearing. Hindu and Buddhist forms are what those bases are understood to be. No source used here prints an azimuth for a named shrine, so an east door toward the sunrise, the usual rule in many Hindu temples, is not given a degree on this ground. A door facing east is a claim you can check with a compass. It has not been checked in the sources used here.
Kota Gelanggi is the other story, and it is in Johor, not in Kedah. In 2005 a published claim placed a lost city of that name in the Linggiu forest and said it might be older than Angkor, from texts and from local accounts. The person who published the claim had not walked the ground. Archaeologists already working in the wider district said settlements older than the fourteenth century, and material from an old fort placed between the ninth and the twelfth, were real, and that a city of walls and a temple could not be confirmed until more fieldwork was done. A museum expedition went. Months later an archaeologist of the heritage department said, in a forum paper, that the team found no evidence of an ancient civilisation in the Linggiu area. The same news report said the official finding had not been released. A field visit that finds no city is evidence against a city. Unreleased paperwork is a reason to want the report, not a reason to invent ruins. Older settlement traces in Johor can be true, and a lost capital older than Angkor can still be absent. Those are different sentences.
No sky reading of Kota Gelanggi exists to weigh, because no plan was published to aim at the sky. The valley’s shrines may yet have one. It is not in hand.
Why it matters
The early history of the peninsula is either an iron port in Kedah with a debated start, or a hidden capital in Johor that has not been stood on. Mixing them makes Kedah’s furnaces carry Johor’s story. They are two arguments, and only one has slag.
The furnace shelter, the restored brick base, and the Johor claim are the tour. The pin is Sungai Batu. Kota Gelanggi has no structure to pin.
Photo tour
2 photographs, unedited. On a phone, swipe the row. Choose one to look closely.
- Kota Gelanggi has no open photograph of a wall or a temple, because the expedition that looked reported none. A picture of forest would pretend to be a city. It is left out on purpose.
What is still unexplained
These sit beside the consensus account, with the same visual weight. None of them is marked solved.
Location and dating
- Location
- Sungai Batu in the Bujang Valley, Merbok, Kedah, between Gunung Jerai and the Muda River. The pin is 5.691 north, 100.457 east. Kota Gelanggi, if the claim’s forest is the right one, is in the Linggiu reserve in Johor, a different state.
- Period label
- Current consensus — open to revision Iron smelting at SB2A in the first to fifth centuries CE on the 2020 study, and mostly second to tenth centuries CE on the 2023 model, with one sample of 788 to 537 BC set aside. The Johor fort material cited in 2005 was placed in the ninth to twelfth centuries. A city older than Angkor was the claim, and it was not given a laboratory year.
- How this dating was done
- Radiocarbon and accelerator mass spectrometry on the furnace site, plus optically stimulated luminescence on buried grains. The 2023 revision is a Bayesian model of those radiocarbon results. Kota Gelanggi’s date was a comparison with Angkor, not a sample.
- What could change it
- More samples in the eighth century BC would force the outlier back into the model. A published plan of the Linggiu expedition, with coordinates and photographs of what was natural rock and what was brick, would let a reader see the empty city rather than take the forum sentence alone.
What is actually there
- SB2A used a bloomery furnace at about 1150 to 1200 degrees Celsius.
- Haematite is traced to Bukit Tupah. Tuyeres and slag are part of the finds.
- The shelter sign says third to fifth century.
- The 2005 Johor claim was made without a visit. A later expedition reported no city.
- Ninth-to-twelfth-century material in an old fort was reported from the wider Johor district.
Open questions
| Question | Status |
|---|---|
| Does one early sample move Sungai Batu back to the eighth century BC? | Open. The 2023 model says not, unless more samples match it. The sample exists. |
| Was there a city at Kota Gelanggi? | Open in the sense that the texts were real enough to send an expedition. The expedition’s reported result was that no ancient civilisation was found. The official report was said to be unreleased. |
The current narrative
Where publication stands today. It is the best-supported reading, and it is open to revision. It is not a verdict.
The current narrative — open to revision
An iron port in Kedah, early centuries CE
The documented early industry is Sungai Batu’s bloomery iron, first to fifth centuries CE, with a later model stretching most samples to the second through tenth, and brick shrines of Hindu-Buddhist type in the same valley. Kota Gelanggi is a claim the ground did not confirm.
What supports this
- Slag, tuyeres, ore from Bukit Tupah, and three dating methods converge on the early centuries CE at SB2A.
- The 2023 model, made from the radiocarbon set, agrees on a long first-millennium industry and sets 788 BC aside pending more samples.
- A heritage-department archaeologist reported no ancient civilisation in the Linggiu forest after the expedition.
What pushes back
- The early sample is real. Calling it an outlier is a modelling choice, and the choice changes if the next samples match it.
- Pre-fourteenth-century settlement and a fort with ninth-to-twelfth-century material were reported in Johor. That is not nothing, even if it is not Angkor.
- The official Linggiu report was described as unreleased, so the forum sentence is the public trace of the fieldwork, not the fieldwork itself.
What would settle it
It is the reading with furnaces, slag, and a pin. It treats the Johor claim as tested and not confirmed, and it leaves the single early radiocarbon sample in view.
Emerging narratives
Other published readings. Each may be right. None is adopted here unless the evidence carries it.
An emerging narrative
A lost city in the Linggiu forest
Kota Gelanggi was a capital in the Johor forest, possibly older than Angkor, known from texts and local accounts, and the failure to map it is a failure of access and of an unfinished report.
What supports this
- A 2005 publication located the city from historical texts and local testimony, and archaeologists in the region had already reported settlement older than the Melaka period.
- Ninth-to-twelfth-century material in an old fort shows that Johor was not empty in those centuries.
- The official expedition report was said not to have been released, which leaves the negative result thinner than a published site plan.
What pushes back
- The claim’s author had not visited the forest. A city located from texts still needs a wall.
- The team that did visit reported no evidence of an ancient civilisation there.
- Nothing in the claim has a radiocarbon age, a plan, or a photograph of masonry. The Kedah furnaces have all three.
What would settle it
It is the reading a reader can still hold if the unreleased report matters more than the forum account of an empty forest. It has no structure to show. The valley does.
Should the current view be reevaluated?
Open to revision
Sungai Batu’s iron industry in the early centuries CE is well supported, and the eighth-century-BC sample is a real loose end rather than a second city. Kota Gelanggi was looked for and not found, on the public account of that search, with the official report still undescribed in the sources used here.
What drives this
- Furnaces dated by radiocarbon, accelerator mass spectrometry, and optically stimulated luminescence.
- One much older sample, later treated as an outlier.
- A 2005 lost-city claim and a field visit that reported no city.
This is the museum’s own weighing of the evidence. It is not a visitor vote, and it is separate from the slider.
Weigh it yourself
Anonymous. Nothing here is stored except on this browser. Your lean does not change the assessment beside it.
Visitor lean so far — not evidence. On this browser only; a shared tally is not collected.
If the contested reading is right
What would follow if a reading that is not the consensus turned out to be true, and what could settle it. This explains the stakes. It does not adopt the reading.
Kota Gelanggi was a capital older than Angkor
The peninsula’s earliest city would stand in the Johor forest, and the Kedah iron port would be a later, or a parallel, coast.
Why it is still worth a test
The expedition report, with coordinates, and any masonry it recorded or explicitly did not find.
The 788 BC sample is the true start
Iron at Sungai Batu would begin in the early first millennium BC, and the brick shrines would be a late chapter of a much longer industry.
Why it is still worth a test
More radiocarbon samples from the same early context. One specimen will not carry the model.
Sources
- The iron smelting technology of the Bujang Valley Archaeopress
- Bujang Valley, revised dating The Vibes
- Johor relics predate Malacca The Star
- Is there a lost city in Linggiu? Malaysiakini
Image credits
- The shelter over the iron-smelting site at Sungai Batu. The sign reads “iron-smelting site, 3rd–5th century,” Sungai Batu, Bujang Valley. Furnace remains sit in the shade. The sign is a label, not a laboratory sheet. . Blackhurrem, CC0 1.0. Shown unedited.
- A restored brick shrine base in the Bujang Valley. The courses are even, and the lawn is modern. No compass bearing is published for this door in the sources used here. . Anandajoti, CC BY-SA 3.0. Shown unedited.