The participation of the author in the ISBSA conference in Marseille was made possible by the Seafaring: Maritime knowledge in the South China Seas project (ANR/École française d’Extrême-Orient, Paris/Academia Sinica, Taipei).

1Southeast Asian shipbuilding processes have by now been sufficiently documented, with data gathered from textual sources, ethnographic observations and, more recently, from archaeological research, to start distinguishing paths of development in technologies used to assemble hulls in the past two millennia. It is now possible to build up a tentative typology of this technical tradition in both chronological and regional terms. In this short contribution, I will concentrate only on how hull planks were assembled and then fastened to frames, in a shell-first shipbuilding process that survived until our days in Indonesian waters. Such defining features are easily observable, even in the most disturbed archaeological contexts.

2The oldest recorded boat remains found in Southeast Asian archaeological sites date from the first half of the first millennium AD. Their hulls were assembled with vegetal fibres, using a technique belonging to a broadly defined “sewn-plank tradition”, but with clear local specificities, usually designated as the “stitched-plank and lashed-lug tradition” (Manguin 1985, 2019; Gibson-Hill 1952; McGrail 2001, p. 297-298; Flecker 2012; Liebner 2014; Lacsina 2015). The Southeast Asian tradition needs to be clearly distinguished from the other major sewnplank tradition of Asian seas, characteristic of the craft of the western Indian Ocean, also known as the “Arabo-Indian tradition” (Fibre and Wood 2019; Staples, Blue 2019). Contrary to the latter tradition, where sewing is done with a continuous thread passed through holes drilled across the planks, therefore visible on the outer side of the hull, the Southeast Asian technique uses discrete lashings, disconnected from each other, that pass through holes drilled into the edge and out of the inner face of planks. Stitches are therefore not visible on the flush outer surface of the hull. Ropes used to assemble these vessels were almost always made of the fibre of the sugar palm tree (Arenga pinnata or A. saccharifera).

3In the earlier ships observed in Southeast Asia (dated to around the 4th to 5th century AD), the fastening of the planks with separate lashings was reinforced by only a few treenails inserted at a distance from each other (some craft may have been built without such dowels, but reliable evidence is lacking). All planks were fastened to the frames (and to the thwarts) by way of protruding lugs or blocks carved out on the inner face of planks (a feature independently attested in Scandinavian ancient traditions, but unknown to the Arabo-Indian or Mediterranean traditions) (fig. 1).

4The craft dated to the second half of the first millennium AD still had their planks stitched together with vegetal fibres, but the number of treenails clearly increased, providing more structural rigidity to the hull. The occasional tenon-and-mortise joint also appears to have been used in places to lock planks together, as observed so far only in the 7th-century Sambirejo (South Sumatra) and Punjulharjo (Central Java) sites (Manguin 2019, fig. 4). On the other hand, the lashing of planks to frames via lugs still remained in general use.

5At the turn of the second millennium AD, a major change occurred. Whereas the lashing of planks to frames via lugs again remained in use, the sewing of planks with discrete lashings disappeared in all known ship remains of Southeast Asian build. Only treenails, by now more numerous, remained to fasten planks together. Treenails alone may not have been sufficient to keep the hull in compression during its construction in a shell-first shipbuilding process, or to prevent the strakes from riding apart when sailing, with only the lashings between frames, thwarts and planks left to keep the hull in place. Driving smaller locking treenails into some of the dowels from the inside of the planks became a common practice to compensate the loss of the lashings, as clearly attested in Southeast Asian craft discovered in archaeological contexts. This technique was still to be observed in purely dowelled plank-built boats in Java in the 1980s (figs 1-3). Another technique that fulfils the same purpose, known among 20th-century shipbuilders of South Sulawesi (Indonesia) but never found in an archaeological context, is to drive a wedge into the head of a treenail (one out of every three or four) left protruding after assembling two strakes (figs 2 and 4) or assembling planks to the frames.

6Whereas the process of forsaking the sewing of the planks appears to have been general in Southeast Asian waters in the second millennium, the lashing of frames to lugs carved out of planks remained attested in the Philippines and the Moluccas for smaller craft in the 16th and 17th centuries, and even survived to this day in whaling boats of the small island of Lembata, in eastern Indonesia (Manguin 2019, p. 402; Barnes 1985).

7So far no sites with remains of ships from the 15th and 16th centuries have been found in Southeast Asia. However, keen Portuguese observers precisely described the large traders they then encountered in the Melaka Strait and the Java Sea, which were built by the major maritime powers of insular Southeast Asia. The Malays and Javanese called them jong, the term at the origin of the modern word junk. These concurring texts tell us that a further major change had taken place in some Southeast Asian shipyards: assembly of the hull was now entirely done by way of treenails, both between planks and between planks and frames (to the total exclusion of iron nails or clamps, as in earlier times) (Manguin 1980, 2019). Lashings with vegetal fibres were no longer in use for assembling these large vessels, and the protruding lugs carved out in the inner side of the planks must have then disappeared. Such fully treenailed ship structures survived in Indonesian waters until late in the 20th century. The fact that Indonesian shipbuilders kept alive the process of locking the treenails and the strakes in place with smaller treenails or with wedges is one good indication that 15th-16th century and later ships are the final step in the evolution from the earlier sewn techniques used in Southeast Asian waters. The first major step in this development was the rejection of stitches between planks and their progressive replacement by treenails, keeping the lashings only to fasten the planks to the frames. These lashings of planks to frames via lugs were in turn discarded and also gave way to treenails to fasten the frames to the hull.

8There is sufficient evidence at hand, therefore, to establish that Southeast Asian shipbuilding techniques underwent significant evolution during the past two millennia. All the available data confirms the continuity between shipbuilding procedures of the first and second millennia AD. It is relevant here to note that a comparable evolution – from sewing to treenails (and/or tenons and mortises), with composite intermediary phases – took place in Greek vessels around the mid-first millennium BC (Pomey 2010). Compared to the vast corpus of Mediterranean ships, however, we still have a very limited amount of data at hand in Southeast Asia, with only some 30 sewn vessels available for study, many of them fragmentary or from insufficiently controlled excavations. These craft are, moreover, spread over two millennia, and dispersed over a vast and culturally very diverse area, within a multitude of mainland and insular Southeast Asian ecosystems, harbouring peoples of a variety of ethnolinguistic groups. Sizes, carrying capacities and functions of these vessels may have widely contrasted. The largest ships were built within societies that progressively formed complex polities of various magnitudes, which played an active role in shaping exchange and trade networks in the Indian Ocean and the South China Sea. Many of the ship remains from South Sumatra were thus discovered during surveys and excavations in the city of Palembang and its surroundings, the capital of the powerful thalassocracy of Srivijaya (7th–13th century) (Manguin 2004). Others may have been used by fishermen belonging to simpler societies, many of which survived in remote parts of insular Southeast Asia well into the 20th century.

9If there is little doubt that Southeast Asian technologies evolved over the centuries along the pattern described above (a pattern shared with other regions in the world), we obviously do not have enough examples to fully describe and understand this evolution. One may speculate that improvements could have been introduced in larger trading ships, where more structural strength was required to carry large cargoes. The large 10th-century Cirebon wreck may thus represent a transition phase. Her upper planks were fastened by dowels alone, with lugs still carved out to lash them to frames and thwarts. The bottom strakes, for lack of clear evidence, are harder to interpret as they were not dismantled during excavation: the holes in the flat (lugless) bottom strakes could have been used to lash the latter to the floors, with planks fastened together with treenails (Liebner 2014). However, the smaller boats excavated in Butuan, dated to between the late 8th and the 10th century, also appear to pertain to this new trend in shipbuilding, providing the earliest example of true lashed-lug only vessels with no more stitches for fastening the planks (Lacsina 2015).

10As was often noted for Mediterranean vessels, such evolution in Southeast Asian seas may not have been unilinear. Not all the vessels examined above would have served the same functions, and interactions between regional variants must have taken place. We will need many more properly studied craft, brought to light during systematic excavations, to better distinguish between internal evolution and technological transfers within or without the Southeast Asian shipbuilding tradition.