Navigation – Plan du site

AccueilNuméros21Inland Ship ConstructionThe Penglai 3 Shipwreck in Contex...

Inland Ship Construction

The Penglai 3 Shipwreck in Context: Korean Shipbuilding in the Late 14th and Early 15th Centuries

Damien Peladan
p. 141-145


Les derniers développements de l’archéologie navale coréenne ont fourni une image plus claire de la construction navale coréenne à l’époque médiévale. Cependant l’épave Penglai 3, fouillée en Chine et datée de la fin du XIVe ou du début du XVe siècle, bien que clairement coréenne dans sa structure de base, présente des caractéristiques qui n’ont jamais été observées ailleurs sur d’autres épaves coréennes. Cet article vise à expliquer le caractère unique de ce navire en examinant le contexte historique spécifique dans lequel il a été construit, à une époque où la cour coréenne poursuivait diverses expériences techniques.

Haut de page

Texte intégral

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/Ecole française d’Extrême-Orient, Paris/Academia Sinica, Taipei).

1For most of its history, Korea has had its own unique ship-building tradition, distinct from the neighbouring countries of Japan and China, a fact which has recently come to light thanks to the development of underwater and intertidal naval archaeology in South Korea. The first ancient shipwreck was excavated in 1984, followed by a second one in 1995. Since then, the pace of discovery has been rapidly increasing, to the point that so far 13 ancient wrecks have been excavated in Korean waters, and two others in the city of Penglai in Shandong province, China. The vast majority of these shipwrecks are dated to the Middle Ages (fig. 1), so we now possess a good overview of what medieval Korean shipbuilding looked like, especially for the Koryo era (918-1392), when maritime activity was at its greatest in the waters of the peninsula.

  • 1 KRS: 104, 10b Kim Panggyong chon (“Biography of Kim Panggyong”), 15th year of Wonjong’s reign (127 (...)

2However, among these 15 wrecks, the Penglai 3 shipwreck (one of the two excavated in Penglai, China, in 2005) stands out from the lot: although typically Korean in its basic structure and building techniques, it exhibits features which have never been observed on wrecks excavated within Korean waters, some of which were clearly influenced by Chinese shipbuilding. Several hypotheses have been proposed to explain these peculiar shapes. For some archaeologists, they are the result of technology transfers from China to Korea at the time of the failed Mongol invasions of Japan (in 1274 and 1281), when the Mongols ordered the newly conquered Korean court to construct an invasion fleet using Chinese shipbuilding techniques (Yuan 2011, p. 54). However, the very same historical source which reports this fact also states that the costs of such an enterprise would have been too high and it would have taken too long, so the Koreans ended up using indigenous techniques1. It is therefore rather doubtful that this event would have been the occasion of considerable technology transfers. Moreover, the Tallido shipwreck, excavated in 1995 and dated to the second half of the 13th or first half of the 14th century does not exhibit any of the features found on the Penglai 3 wreck. Another explanation is that the Penglai 3 was in fact an oceangoing ship, designed to cross the sea from Korea to China, while the wrecks found within Korean waters were coasters, which would explain its singularities (Kim 2009, p. 265-266). However, a few metres away from the Penglai 3, the remains of another Korean wreck were also excavated, the Penglai 4, which does not exhibit any of the characteristics of the Penglai 3, somewhat undermining this hypothesis. Moreover some of these special features have also been found on contemporaneous wrecks in Korean waters, as we shall discuss below.

3The present paper thus aims to explain the uniqueness of the Penglai 3 shipwreck through its specific historical context, as it was built at a time when the Korean court, facing many challenges from across the sea, was experimenting with various techniques in order to improve its own navy.

Fig. 1: Ancient Korean shipwrecks excavated from 1983 to 2019

Fig. 1: Ancient Korean shipwrecks excavated from 1983 to 2019

1. Overview of Medieval Korean Shipwrecks and Shipbuilding

  • 2 The only exception is the Taebudo 2 shipwreck, which has four bottom planks. All the other wrecks (...)

4First and foremost, let us begin with an overview of Korean shipbuilding in the Koryo era. Traditional Korean ships were entirely made of wood (mostly pinewood) and had the particularity of not using any iron nails or clamps in their structure. The components of the hull were joined together using grooved clinker joints, mortise-and-tenon joints, wooden pegs and wedges. They were flat bottomed so as to navigate safely the coastal waters of the peninsula, strewn with mudflats and intertidal zones. The bottom was made of an odd number of rows of thick planks (usually 3, 5 or in some cases 7)2, which were held together using a long wooden peg (called changsak) going through the whole width of the bottom planking. In earlier wrecks, these bottom planks tended to be very thick, and then got thinner over time.

5The side strakes were attached on either side of the bottom planking, rising at a steep angle, and joined together using an elaborate system of grooved clinker joints and wooden pegs (called p’isak). The strakes on both sides were then held together using throughbeams arranged in rows, which would act as a frame. The way these strakes were connected to the bottom planks evolved over time. While earlier ships used one or two rows of L-shaped chine strakes on either side to serve as a transition, these seem to have been completely abandoned in the early 13th century. From then on, the bottom side strakes were directly attached to the bottom planking using a groove joint. Korean shipwrights would typically waterproof the hull by tightly packing rice straw rope in between the planks.

  • 3 Xu Jing (1091-1153), Xuanhe fengshi Gaoli tujing (“Illustrated Account of an Official Mission to K (...)

6All of the wrecks dated from the 11th to 14th centuries had only one mast, as evidenced by the two slots carved in the centre of the middle bottom plank, where the mast support was directly inserted, without any intermediary mast step. Textual evidence from the early 12th century confirms that it was the norm for Korean ships to have only one mast3. It was placed forward of the central row of throughbeams, against which it directly rested (L. Chul-Han, S. Randall 2010).

7The size of the ships varied according to their use, from about 9 m for the smaller (Wando shipwreck) to 30 to 40 m long in the case of the Naju shipwreck, although only a few bottom and side strakes have been found. The length of most of the ships ranged from 12 to 17 m (Mun 2014, p. 171-184) (fig. 2).

Fig. 2: Reconstruction of Mado 1 shipwreck (sank in 1208)

Fig. 2: Reconstruction of Mado 1 shipwreck (sank in 1208)

This class of ship probably corresponded to ch’omasŏn (government- operated tax transport ships).

(from NRIMCH 2015, p. 118)

2. The Characteristics of the Penglai 3 Shipwreck

8As stated above, the Penglai 3 shipwreck is the most peculiar Korean shipwreck excavated to date. Dating from the late 14th to early 15th century (but no later than 1420) (Kim 2009, p. 259-260), it is rather large in comparison to most Korean wrecks (the ship’s reconstructed length is estimated to be about 22 m). As is the case of other Korean ships, it was primarily made of pinewood. Flat bottomed, its side strakes were fastened using grooved clinker joints and wooden pegs, and its overall shape and basic structure can be qualified as typically Korean. On the other hand, it used bulkheads instead of through- beams in order to strengthen the hull. Furthermore, a plank resembling a keelson was attached on top of the middle row of bottom planks, and iron nails were also used along with the wooden pegs in order to reinforce the structure of the hull. It also had two masts, as evidenced by the presence of two mast steps, another innovation never found on other Korean ships, and it used a lime-based compound as luting material, as opposed to typical rice straw rope waterproofing (fig. 3-4).

Fig. 3: Remains of the Penglai 3 shipwreck

Fig. 3: Remains of the Penglai 3 shipwreck

(from CR&AISP 2006, p. 36)

Fig. 4: Reconstruction of the Penglai 3 shipwreck

Fig. 4: Reconstruction of the Penglai 3 shipwreck

(from CR&AISP 2006, p. 118)

9Most of these features can be traced back to either Chinese or Japanese shipbuilding. The use of bulkheads for instance is ubiquitous in Chinese shipbuilding, as is the use of keelsons (which is not surprising as Chinese seagoing junks possessed a V-shaped hull constructed around a keel), mast steps and lime-based caulking compound. However, Chinese caulking compound mixed tung oil with the lime as a binding agent, while the compound found on the Penglai 3 ship solely used lime and was, as a result, of poorer quality (Yuan 2011, p. 52-53). The use of iron nails was also common in both Chinese and Japanese shipbuilding.

10The Penglai 3 can be compared to two other wrecks dating from the same period found in Korea. The Anjwa shipwreck, dated to the late 14th century, is one of the largest Korean ships excavated to date, with an estimated total length of 17 m. The cross section of the hull shows a flat bottom with a gentle wine- glass-like turn of the bilge, slightly more elegant than earlier shipwrecks, with similarities to the Penglai 3. Although it does not have any mast step, the two mast slots on the middle bottom plank were dug into a slightly elevated platform carved on the plank, to which it added an extra third in width, a feature that so far has only been observed on this shipwreck. Another interesting point of the Anjwa ship is that it used, in addition to the traditional rice straw rope stuffed in between the planks as waterproofing, a lime-based luting material similar to the one found on the Penglai 3 (NMM 2006, p.  32, 81, 88 and 96). In fact, historical records show that Koreans knew about using lime as a luting material at least since 1412 (CWS: 1412.5.21 [3]), but only learnt about the Chinese technique of mixing tung oil (or, when not available, fish oil or sesame oil) with the lime in 1451 (CWS: 1451.3.3 [2]).

11Another contemporaneous point of comparison for the Penglai 3 is the Mado 4 shipwreck (NRIMCH 2016). Excavated in 2015, it was a government operated tax transport ship which sank en route while sailing from the south to the capital, sometime between 1417 and 1425. It is the most recent medieval wreck to be found, dating from roughly the same period as the Penglai 3. It too has a few noteworthy features, such as the sparing use of iron nails, as well as two sets of mast slots on the middle row of bottom planks, indicative of a second mast, something which had never been witnessed on earlier wrecks found in Korean waters. As the wreck has not yet been dismantled for study, we do not know what type of caulking was used.

12In summary, even though most of the features found on the Penglai 3 shipwreck can be traced back to Chinese and Japanese shipbuilding, some of them are also present on other Korean shipwrecks of the same time period, hinting at the fact that these innovations were not restricted to the Penglai 3, nor were they necessarily particular to oceangoing ships, but were in fact the result of a broader current within Korean society. It is thus essential, in order to explain the characteristics of the Penglai 3, to look deeper into the historical background of the late Koryo and early Chosŏn era.

3. Evolution of Shipbuilding in the Late 14th and Early 15th Century

13In the second half of the 14th century, the Korean peninsula was faced with an unprecedented challenge coming from across the seas: Japanese pirate fleets, gathering sometimes up to several hundred ships, raided the Korean coast year after year, rampaging and plundering villages, temples, towns, military outposts and of course ships, and capturing the populations to be sold as slaves on Japanese markets. The Koreans soon realized that their navy was ill equipped to fend off these attacks and started building new warships from the 1370s and 1380s onwards, progressively taking back control of their waters. These efforts were carried on into the new Chosŏn dynasty (1392-1910). In this process of rebuilding its naval might, the Korean court started paying a great deal of attention to Chinese and Japanese shipbuilding techniques and comparing them to their own, especially during the first half of the 15th century. They would compare the structure of the hull, the materials used for the sails and anchors (CWS: 1446.9.9 [2]), the types of wood (CWS: 1430.5.24 [1]), etc. However, the curiosity for foreign shipbuilding techniques did not stop at mere observation. Indeed, Korean authorities invited Japanese shipwrights to Korea, and had them build ships there (CWS: 1417.5.19 [2]). These ships were not meant to replace indigenous shipbuilding, but rather were used to conduct various experiments. For example, in 1413, the king ordered the conduct of speed comparisons between Japanese and Korean ships. As it turned out, the Japanese ship outsailed its Korean counterpart by 30 to 40 paces when sailing with the current, and over 100 paces when going against (CWS: 1413.1.14 [1]). The Korean authorities thus paid a great deal of attention to the strengths and weaknesses of their own ships compared to foreign vessels and tried to bridge the gap whenever possible. They also used them for experiments in naval warfare, and the development of techniques to counter Japanese tactics, such as the creation in 1415 of “turtle ships” (kwison), warships whose upper deck was covered with a spiked roof to serve as protection against projectiles and to prevent the enemy from boarding (CWS: 1415.7.16 [2]). In other instances, Japanese ships built in Korea were also used as target practice for the newly developed cannons (CWS: 1445.9.22 [2]).

14Scholars who have studied the Penglai 3 usually assume that the iron nails were introduced to Korea from China, but one cannot rule out the influence of Japanese shipbuilding in this regard. For instance, in 1421, local authorities in Kyongsang Province asked that the 34 Japanese ships captured in Japan two years prior be dismantled and their iron nails taken out to be used for the construction of new warships (CWS: 1421.3.14 [7]). Japanese nails were indeed highly praised by the Koreans, as evidenced by several texts. It is therefore quite possible that the introduction of such nails as seen on both the Penglai 3 and the Mado 4 were, in fact, the result of experimenting with Japanese shipbuilding techniques in the early Chosŏn period.

15Evidently, however, Japanese shipbuilding was by no means the only source of inspiration for Korean shipwrights. Chinese techniques, being the most advanced in East Asia at the time, were also the focus of Korean enquiries and experiments. This trend can be traced back to the year 1415, when the king mandated two Chinese men, Liu Siyi and Li Xuan, to conduct experiments with Chinese shipbuilding, because “the military ships in each province before long get eaten by worms and have to be rebuilt after a few years. Not only is it taxing on the workforce, but [Korean] timber resources cannot keep up” (CWS: 1415.11.19 [4]). It is interesting to note that this event occurred roughly around the same time as the Penglai 3 was built. One cannot but wonder if this ship was not, in fact, the direct product of those experiments.

  • 4 Interestingly enough, the Koreans tried to build a double-planked hull during the initial construc (...)

16It is quite clear that one of the aspects of Chinese shipbuilding which appealed the most to Korean authorities was durability. The rebuilding of the navy had almost depleted the timber resources of the peninsula, to the point that the Chosŏn court had to create state forests all along the coast in 1448 (CWS: 1448.8.27[1]). Chinese ships had the reputation of lasting much longer than their Korean counterparts, in part due to the use of double planking, a technique with which the Korean court also experimented in the 1430s and 1440s (CWS: 1430.5.24 [1]). This particular experiment was eventually discontinued in 1451 since it was not producing the desired effect, while being more costly than single-layer planking (CWS: 1451.6.12 [4])4.

4. Conclusion

17In conclusion, the peculiar appearance of the Penglai 3 shipwreck cannot be separated from the specific historical context in which it was built. It was not the result of technology transfers in the late 13th century nor a typical oceangoing Korean vessel, as much as it was the product of the numerous experiments conducted by Korean authorities in the early 15th century, as evidenced by the fact that some of its unusual features have also been found on other contemporaneous ships excavated in Korean waters.

18The enthusiasm of Korean authorities for such innovation and experimentation started to wane in the second half of the 15th century. Many of the techniques newly introduced, such as double planking for the hull, iron anchors, and indeed most of the hybrid features found on the Penglai 3, seem to have been abandoned thereafter. Some of them, however, were retained by Korean shipwrights, and remained an integral part of Korean shipbuilding until the early 20th century. Indeed, although no Korean wreck later than the early 15th century has been discovered so far, we know from pictorial and textual evidence from the 18th and 19th centuries that some of the features adopted in the early Chosŏn period became the norm, such as the use of iron nails as a supplement to wooden pegs (NRIMCH 2016, p. 388), and the construction of two-masted ships, which seem to have become common for middle-sized and larger ships for most of the Chosŏn period, from fishing vessels to warships (NRIMCH 2012, p. 139-158). In this respect, it seems fair to say that the late 14th and early 15th centuries were a defining period for Korean shipbuilding.

List of Abbreviations

19CR&AISP: Cultural Relics & Archaeological Institute of Shandong Province, based in the city of Penglai, China.

20CWS: Chŏson Wangjo Sillok (Veritable Records of the Chŏson Dynasty).

21KRS: Koryŏsa (History of Koryo), compiled in 1451. NMM: National Maritime Museum, present day National

22Research Institute of Maritime Cultural Heritage (see below).
NRIMCH: National Research Institute of Maritime Cultural

23Heritage, based in the city of Mokp’o (South Korea).

Haut de page


Cultural Relics & Archaeological Institute of Shandong Province
2006 Penglai guchuan (“Ancient ships from Penglai”), Beijing, Wenwu Chubanshe.

Kim Sŏngbŏm
2009 Chungguk Pongnae susong ch’ult’o Koryoson (“Koryŏ shipwrecks excavated in the Penglai Shuicheng in China”), in Hanguk Chungsesa Yŏn’gu 27, p. 233-284.

Mun Kyŏngho
2014 Koryŏ sidae chounjedo yŏn’gu (“Research on the Koryo Era Tax Transport System”), Seoul, Hyean.

National Maritime Museum
2006 Anjwasŏn palgul pogosŏ (“Excavation report of the Anjwa shipwreck”), Mokp’o.

National Research Institute of Maritime Cultural Heritage
2012 Chounsŏn – Chŏnt’ong sŏnbak chosŏn kisul (“Tax transport ships – Traditional shipbuilding techniques”), Mokp’o.

National Research Institute of Maritime Cultural Heritage
2015 Koryŏ sidae chounsŏn (Mado 1hosŏn) pog’wŏn pogosŏ (“Report on the Reconstruction of a Tax Transport Ship (Mado n°1) of the Koryo Era”), Mokp’o.

National Research Institute of Maritime Cultural Heritage
2016 T’aean Mado 4hosŏn sujung palgul chosa pogosŏ (“The Mado 4 shipwreck in T’aean – Underwater excavation report”), Mokp’o.

Yuan Xiaochun
2011 Pongnae Koryŏ (Chosŏn) kosŏn-ŭi kujo-wa t’ŭkching (“Structure and specificities of the Penglai Korean ships”), in National Research Institute of Maritime Cultural Heritage (ed.), Koryŏ-ui nanp’asŏn-gwa munhwasa (“Koryo Shipwrecks and Cultural History”), Mokp’o, Research Centre for Island Culture, p. 45-65.

Lee Chul-Han, Sasaki Randall
2010 Goryeo Dynasty (918-1392) shipwrecks in Korea, in Kimura Jun (ed.), Shipwreck ASIA: Thematic Studies in East Asian Maritime Archaeology, Adelaide, Shipwreck ASIA, p. 56-73.

Haut de page


1 KRS: 104, 10b Kim Panggyong chon (“Biography of Kim Panggyong”), 15th year of Wonjong’s reign (1274).

2 The only exception is the Taebudo 2 shipwreck, which has four bottom planks. All the other wrecks have an odd number.

3 Xu Jing (1091-1153), Xuanhe fengshi Gaoli tujing (“Illustrated Account of an Official Mission to Korea during the Xuanhe Reign”), Book 33, Xunchuan (“Patrol ships”); Guanchuan (“Government ships”).

4 Interestingly enough, the Koreans tried to build a double-planked hull during the initial construction of the ship (see for example CWS: 1451.5.25 [3]), while Chinese shipwrights tended to add outer layers of planking later in the life of the ship, as the first layer became worn.

Haut de page

Table des illustrations

Titre Fig. 1: Ancient Korean shipwrecks excavated from 1983 to 2019
Fichier image/jpeg, 95k
Titre Fig. 2: Reconstruction of Mado 1 shipwreck (sank in 1208)
Légende This class of ship probably corresponded to ch’omasŏn (government- operated tax transport ships).
Crédits (from NRIMCH 2015, p. 118)
Fichier image/jpeg, 380k
Titre Fig. 3: Remains of the Penglai 3 shipwreck
Crédits (from CR&AISP 2006, p. 36)
Fichier image/jpeg, 305k
Titre Fig. 4: Reconstruction of the Penglai 3 shipwreck
Crédits (from CR&AISP 2006, p. 118)
Fichier image/jpeg, 314k
Haut de page

Pour citer cet article

Référence papier

Damien Peladan, « The Penglai 3 Shipwreck in Context: Korean Shipbuilding in the Late 14th and Early 15th Centuries »Archaeonautica, 21 | 2021, 141-145.

Référence électronique

Damien Peladan, « The Penglai 3 Shipwreck in Context: Korean Shipbuilding in the Late 14th and Early 15th Centuries »Archaeonautica [En ligne], 21 | 2021, mis en ligne le 01 juillet 2022, consulté le 08 décembre 2023. URL : ; DOI :

Haut de page


Damien Peladan

Université de Paris, Paris, France

Haut de page

Droits d’auteur


Le texte seul est utilisable sous licence CC BY 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search