The ‘lost city’ of Kota Gelanggi: an exploratory essay based on textual evidence and an excursion into ‘aerial archaeology’
Raimy Ché-Ross’s 2004 exploratory essay, published in the Journal of the Malaysian Branch of the Royal Asiatic Society, re-examines the long-standing identification of the legendary “lost city” of Kota Gelanggi with the cave complex of the same name in Pahang, arguing instead that the city described in the Sejarah Melayu lies on the upper reaches of the Johor River and presenting preliminary aerial and ground evidence for a candidate site in central Johor.
Summary
The article addresses a historiographical impasse that has persisted since the 1930s: whether the “Kota Gelanggi” (or “Kota Batu Hitam”) mentioned in Chapter II of the Sejarah Melayu—described as a great fortified city of black stone, raided by the Chola king Rajendra Cholavarman I around AD 1025—corresponds to the limestone cave complex near Jerantut in Pahang, as Linehan (1936) and Swettenham (1885) had proposed, or to an as-yet-undiscovered site in Johor. Ché-Ross dismantles the Pahang identification on multiple grounds: the Sejarah Melayu explicitly describes a “large city built of stone as black as peppercorns,” which is incompatible with a cave system; geological surveys confirm that the Hulu Sungai Johor region is underlain by granite and hornblende granite in shades of bluish-black to grey-black, providing a plausible building material; and no archaeological excavation at the Pahang caves has ever yielded evidence of habitation by an advanced culture postdating the Neolithic period. The author further argues that the name “Gelanggi” derives from the Thai “Khlang-Kaew” (Treasury of Jewels), a reading consistent with the Sejarah Melayu’s own etymological gloss, and that the city’s location on the Johor River is corroborated by the emphatic and repeated assertion in later recensions that “sekarang lagi ada kota itu di hulu Sungai Johor.”
A second thread of the argument concerns the identity of the “Goa” referenced in the Muqadimah of the Sejarah Melayu, which has been variously interpreted as a cave in Pahang or the Portuguese port of Goa in India. Ché-Ross conducts a systematic concordance search of the full Recension II text and demonstrates that the word “Goa” appears thirty times—nineteen in reference to the Bugis city of Goa in Macassar and eleven in reference to Portuguese Goa—never as a cave. The word “Pahang” appears 268 times in the same text without any association with a “Gua,” and no contemporaneous Malay work (including the Hikayat Hang Tuah and the Bustanus-Salatin) links Pahang to a cave. The author concludes that the nucleus text of the Sejarah Melayu was brought to Johor from Bugis Goa in Macassar, a reading that aligns with the work of Douglas, Roolvink, and Abdul Samad Ahmad.
The article’s final section reports on a private aerial reconnaissance and ground expedition conducted in September 2003 in the vicinity of the Sungai Linggiu in central Johor. Aerial photographs revealed geometric anomalies including what appears to be a “boxed” set of double walls—a large rectangle enclosing a smaller one—and a grid-like network of pathways converging on a central open square. The ground team encountered deep ditches, earthen embankments, and a square platform framed by parallel mounds, though the expedition was curtailed before the “double walls” could be directly inspected. Ché-Ross frames these findings cautiously, noting that the Johor River was likely an arm of the sea at the time Kota Gelanggi flourished (c. AD 900–1200), making the site’s distance from the present coastline less problematic than it appears.
Key Findings
- The Sejarah Melayu (Raffles Malay 18, the earliest extant version) describes Kota Gelanggi as a “large city, built of stone as black as peppercorns” on the upper reaches of the Johor River, with the name “Gelanggi” glossed as “Perbendaharaan Permata” (Treasury of Jewels) (pp. 29–30).
- Archaeological excavations at the Pahang cave complex (Adi Hj. Taha, 1983, at Kota Balai; Kamaruzaman & Zulkifli, 1997) recovered over 3,000 pieces of freshwater shells, pottery, and stone tools attributable to Hoabinhian and Neolithic cultures, with no evidence of habitation by any advanced culture postdating that period (p. 36).
- Geological maps of Johor (Ingham, 1948; Alexander, 1958; Jamaludin, 1992) confirm that the Hulu Sungai Johor locality is characterized by granite and hornblende granite of post-Trias age, with samples displaying colours from deep bluish-black to grey-black (p. 36).
- A concordance search of Recension II of the Sejarah Melayu yields 30 occurrences of “Goa”: 19 referring to Bugis Goa in Macassar, 11 to Portuguese Goa in India, and zero to a cave; “Pahang” appears 268 times with no association to a “Gua” (pp. 39–42).
- Aerial photography over central Johor (September 2003) detected a “boxed” set of clearly defined double walls (a large rectangle enclosing a smaller one) and a grid-like network of pathways converging on a central open square; ground reconnaissance confirmed deep ditches, earthen embankments, and a square platform framed by parallel mounds (pp. 50–52, Figs. 2–7).
- The Chola raid on Kota Gelanggi is dated to c. AD 1025 under Rajendra Cholavarman I (r. AD 1014–44), and the city is argued to have flourished contemporaneously with the Hindu citadel at Lembah Bujang in Kedah, i.e. c. AD 900–1200 (pp. 28, 50).
Conclusion
Ché-Ross’s definitive takeaway is that the identification of Kota Gelanggi with the Pahang cave complex should be “conclusively discarded” and the original premise that the city lies in Johor “firmly reinstated” (p. 39). The aerial and ground evidence, while not yet confirmatory, tips the balance in favour of a multi-disciplinary reassessment of the site, and the author calls for future exploration to be conducted under internationally regulated heritage preservation standards. Should the city be found, he argues, existing ideas of the region’s pre-Malaccan historical mien “would have to submit itself to a radical re-evaluation” (p. 52).
Context
- The article draws on the Raffles Malay 18 manuscript of the Sejarah Melayu held at the Royal Asiatic Society, London, and the DBP MS 86a (Recension II) held by the Dewan Bahasa dan Pustaka, Kuala Lumpur, alongside geological survey maps and aerial photographic material obtained through private fly-overs.
- Historiographically, the essay intervenes in a debate dominated by colonial-era scholars (Linehan, Quaritch-Wales, Winstedt, Swettenham) and challenges the consensus that had calcified around the Pahang identification since the 1930s, while also correcting Wolters’ (1970) symbolic reading of the toponym “Gelanggi.”