Excavations in Kedah and Province Wellesley, 1957
Michael Sullivan’s 1958 report on the University of Malaya Archaeological Society’s 1957 excavation campaign documents work at four sites in southeast Kedah and northern Province Wellesley, extending the known spatial and chronological range of the ancient Indianised city-state of Kataha. The article argues that the religious landscape of this entrepôt was more complex than previously recognised, encompassing not only Hindu and Buddhist shrines but possibly Jain funerary structures, and that the settlement’s southward extension along the Merbok and Muda river systems was more substantial than Quaritch Wales’s 1936–37 survey had indicated.
Summary
Sullivan’s report addresses a fundamental problem in the archaeology of north-west Malaya: the precise location, extent, and religious character of the ancient city-state identified in Chinese and Tamil sources as Kataha or Langkasuka. Building on the foundational survey of Quaritch Wales (1936–37), which had examined twenty-nine sites and established a tentative brick-size chronology, the 1957 campaign targeted the southward extension of the settlement beyond the Merbok Estuary—an area that in early historical times would have consisted of islands, mangrove swamps, and sandy permatangs rather than the continuous coastal plain visible today. The excavations at Matang Pasir, Tikam Batu, Bukit Batu Lintang, and Kota Aur were designed to test whether the city-state’s footprint extended into this zone and to recover datable material, particularly Chinese export ceramics, which Sullivan identifies as the only reliable chronological indicator at these sites.
The core argument emerges from the typological analysis of the structures themselves. At Matang Pasir, Sullivan demonstrates that the rectangular plan, west-facing orientation, surrounding processional path, and stupa-core block distinguish the shrine as Buddhist rather than Hindu—a significant correction, since all previously excavated shrines in the Kedah complex had been identified as Hindu. The bronze Buddha head and lotus flower (presumably from an Avalokitesvara or Tara image) reinforce this identification and point to a Mahayana Buddhist revival in the ninth or tenth century under Sailendra influence. At Tikam Batu, the large carved sandstone pedestal is reclassified from Wales’s suggestion of a fire altar or statue base to the vedika of a lingam, on the basis of its close formal resemblance to Cham pedestals of the seventh to ninth centuries. The site as a whole, however, is dated later (tenth to fourteenth centuries) on ceramic evidence, suggesting the pedestal may have been moved or that the complex was reused over a long period.
The most speculative but potentially significant finding comes from Kota Aur, where two small rectangular brick structures of a type unknown in Malayan or Indian temple architecture were excavated. The discovery of a crushed iron vessel containing perished organic material (analysed as consistent with foodstuffs rather than bone ash) as a foundation deposit, combined with the structures’ elongated plan and east-west orientation, leads Sullivan to tentatively identify them as tombs—possibly of Jain patriarchs, drawing an analogy with the tombs at Moodbidri in South India. This identification, if confirmed, would introduce a new religious dimension to the Kedah complex and would be the first such find in the region.
Key Findings
- The Matang Pasir shrine measures 18 ft 6 in (E–W) by 11 ft 9 in (N–S) internally, with walls 4 ft 6 in thick faced in laterite blocks averaging 12” × 6” × 4”; it is identified as a Buddhist shrine of the ninth or tenth century based on its rectangular plan, processional path, stupa-core block (2 ft 5 in × 2 ft 7 in, 22½ in high), and associated bronze finds (Buddha head, 11⅛ in high; lotus flower, 2½ in long) (pp. 189–192).
- The Tikam Batu east-west wall foundation is over 60 feet long in surviving remains and must have extended to at least 96 feet; bricks average 12” × 7” × 2½” and laterite blocks 11” × 6” × 3”; the site is dated between the tenth and fourteenth centuries on the basis of Sung–Yuan celadon and stoneware found at the wall base (pp. 198–202).
- The Tikam Batu stone pedestal measures 4 ft 2 in square and 1 ft 7½ in deep, carved from fine-grained sandstone with elaborate mouldings; Sullivan identifies it as the vedika of a lingam, comparing it to Cham examples from Mison and Trakieu dated seventh to ninth centuries (pp. 199–200).
- At Bukit Batu Lintang, a granite Dvarapala relief panel 26 inches high was recovered in situ, along with a ch’ing-pai porcelain box (diameter 1¾ in, Yuan Dynasty), a celadon bowl (diameter 5¾ in, late Sung–Yuan), and fragments of sculpture in the Pallava style (seventh to ninth centuries); the occupation level is dated to the thirteenth and fourteenth centuries (pp. 206–212).
- The Kota Aur structures measure approximately 10½ ft long by 4½ ft wide each, built of brick (averaging 10–11” × 6–7” × 2–2½”) on a single course of laterite; a Ming blue-and-white bowl (early sixteenth century) was found above the brick courses, and the iron vessel deposit in structure A contained iron, calcium, phosphate, and traces of zinc consistent with perished foodstuffs (pp. 213–215).
- The Tikam Batu hill had been systematically dynamited by the Public Works Department to provide stone for the approaches to the new Muda River bridge, destroying the majority of the site before the 1957 excavation could begin (p. 197).
Conclusion
Sullivan’s definitive takeaway is that the ancient city-state of Kataha was considerably larger and religiously more diverse than the Quaritch Wales corpus had suggested. The southward extension of the settlement into the Merbok and Muda valleys is confirmed, the presence of a significant Buddhist component is established for the first time through architectural and artefactual evidence, and the possibility of Jain funerary practice is raised as a new line of inquiry. The article also serves as a urgent conservation warning: the destruction of Tikam Batu by road-building works demonstrates that without active protection, the remaining evidence for this major pre-Islamic urban centre will be lost irrecoverably.
Context
- The study draws on the excavation records of Quaritch Wales (1936–37, 1941) and I.H.N. Evans (1925) as its primary comparative corpus, and on the 1955–56 aerial and ground surveys conducted by Sullivan and the Tregonning–de Jong team as its immediate precursor.
- Historiographically, the article corrects the identification of Langkasuka (relocated to the east coast by Wheatley) and reasserts Kedah as Kataha/Kidaram following Braddell and Nilakanta Sastri, while introducing the first architectural evidence for Buddhism and the first tentative case for Jainism in the Kedah complex.