The Early Ryukyu Kingdom (AD 1429 to 1609)
The Early Ryukyu Kingdom (AD 1429 to 1609)
Abstract and Keywords
This chapter studies the archaeology of the Ryukyu Kingdom from the time of the defeat of its rivals in the early fifteenth century AD to its subjugation by Satsuma in AD 1609. In this period the newly unified Ryukyu Kingdom, consisting of Chūzan and its defeated rivals of Sanhoku and Sannan, reached its full development. It interacted on several levels with communities of all types, from royal courts to small traders and pirates, throughout East and Southeast Asia. Within the Ryukyus a distinctive high culture was created. While old connections with the Japanese main islands took on a new form, an entirely new status as tributary state was consolidated with China.
This chapter is concerned with the archaeology of the Ryukyu Kingdom from the time of the defeat of its rivals in the early fifteenth century AD to its subjugation by Satsuma in AD 1609. It covers the archaeology of the capital, Shuri Gusuku, the port town of Naha, and the expansion of the kingdom to the surrounding islands of the Ryukyus.
In this period the newly unified Ryukyu Kingdom, consisting of Chūzan and its defeated rivals of Sanhoku and Sannan, reached its full development. It interacted on several levels with communities of all types, from royal courts to small traders and pirates, throughout East and Southeast Asia. Within the Ryukyus a distinctive high culture was created. While old connections with the Japanese main islands took on a new form, an entirely new status as tributary state was consolidated with China. In turn, the kingdom established its own internal tributary system by subjugating nearby islands and extracting tribute. The building of a royal capital, linked to a substantial port for transshipment, created a unique islandscape within East Asia.
Asato Susumu (2006) has concluded that by the thirteenth century AD, the rulers of Urasoe already controlled central and southern Okinawajima. In the fourteenth century AD their power was contested by Sanhoku in the north, with its capital at Nakijin, and Sannan in the south, with its capital at Ōzato, as described in Chapter 7. Spatial analysis of gusuku sites shows three clusters, which coincide with the historically known political units (Ladefoged and Pearson 2000). The castles within each of the three political units form subgroupings that often coincide with historically known political factions. This would suggest that at one time the three polities consisted of several smaller political factions (p.235) that go back to the Late Shellmound Period (300 BC to AD 1050) as proposed by Asato (1990).
Two types of historical records allow us to reconstruct some of the struggles and victories that took place on sites that are now being excavated. There are local histories, surviving versions of which date from the seventeenth century AD, and official dynastic annals of China and Korea, which record diplomatic missions and sometimes describe the nature of political units. The diplomatic documents collected in Rekidai hōan, mentioned in Chapter 8, are dated from AD 1424 to 1867 (Okinawa Kenritsu Toshokan 1992). Murai (2000) states that the only original copy of this huge set of documents was destroyed in the Kanto Earthquake in Tokyo and the Battle of Okinawa, and that the surviving photocopies are not clear in some parts, and some passages are difficult to interpret because of unusual Chinese character forms. Beginning in the late fourteenth century AD, there are also Chinese and Korean official histories (Chinese zhengshi, Korean chōngsa) and veritable records (Chinese shilu, Korean sillok) of the Ming, Qing, Koryo, and Chōson; correspondence with Japan’s Shimazu family and the Muromachi Bakufu (military government); historical sources pertaining to Japan’s Zen sects; the Omoro soshi songs of the Shuri court; and official warrants or licenses of the Ryukyu Kingdom (jireisho) (Murai 2008, v).
In the thirteenth century AD parts of Okinawa Island were controlled by King Eiso from his political center of Urasoe. In the fourteenth century AD, competing polities in the north, central, and southern regions of Okinawa Island controlled regional confederacies. Unification under the central Chūzan Kingdom with its capital at Shuri took place in 1429 AD. The victorious Ryukyu Kingdom compiled the Chūzan seikan as its history in the seventeenth century AD, outlining a succession of three early dynasties (Haneji 1983). These were the Tenson Dynasty, the Eiso Dynasty (AD 1260 to 1349), and the Satto Dynasty (AD 1350 to 1405) (Appendix 2). From its beginning in legend, the first (Tenson) dynasty is said to have continued for twenty-five generations, finally to be defeated by a local lord, Shunten, said to be the son of the daughter of the Lord of Ōzato and a Japanese prince, Minamoto Tametomo, who drifted to the Ryukyus after the defeat of the Minamoto Clan by the Taira Clan in the late twelfth century AD (Kerr 1958, 58). The link to Japanese rulers, through Minamoto Tametomo, a descendant of the Japanese emperor Seiwa (reigned AD 858–876), is thought by some scholars to be a creation of the seventeenth century AD, to rationalize the Japanese subjugation of the Ryukyus after AD 1609. Kerr notes that the author of the Chūzan seikan, Haneji Chōshū, was “a regent whose policy centered in the need to reconcile and accommodate Okinawan interests and interests of the Japanese” (Kerr 1958, 46). In addition to this early history of the Chūzan ancestors, other sources give glimpses of local rulers (p.236) in the north (Nakijin) and the south (Ōzato) as well as lesser centers (Pearson 2001, 248–249).
In the late fourteenth century AD all three powerful polities sent diplomatic missions to China, and Chūzan and Sannan established relations with Korea as well. These are recorded in the Chinese and Korean official histories. Following the collapse of the Chinese Yuan dynasty in AD 1367, the first Ming emperor, Hongwu, sent missions throughout East and Southeast Asia to proclaim the dynasty’s legitimacy. Chūzan, Sanhoku, and Sannan each sent their diplomatic missions, with Satto, ruler of Urasoe, being the first Okinawan ruler to be recorded in the Chinese records as a tribute sender (Takara 1993, 53). After him all of the Chūzan rulers are mentioned in the Ming dynasty records with the exception of Shō Sen-i, who ruled for only a year in 1477 AD. Rulers before Satto are not mentioned in the Chinese records of either the Song or Ming dynasties, casting some doubt on early Okinawan documents such as the Chūzan seikan (which were not written as modern historical chronicles). The period of the three contending polities lasted from the early fourteenth century until AD 1422, according to Wada (2006), or 1429 according to Takara (1987, 15).
There was also an important increase in agricultural productivity from the time of the beginning of the First Shō Dynasty, around the beginning of the fifteenth century AD, when two crops of rice were planted annually (Asato Susumu 1998, 58). This innovation came from Fujian with the early tribute missions and was restricted to Okinawajima, according to Korean castaways’ accounts. In the Amami and Sakishima Islands two harvests were obtained from one planting, the plants being semiperennial.
Following the unification, there was a period of very active commerce and international contacts, coupled with a rapid succession of rulers. After the death of Shō Hashi in AD 1439, from 1440 to 1469, there were five rulers in twenty-nine years and two power struggles in the 1450s. This period of instability contrasts with the succeeding fifty-year reign of Shō Shin, from AD 1471 to 1526. Takara describes this as a period of consolidation of the kingdom, in which Shō Shin achieved increased centralization and regional control (Takara 1987, 17).
In the 1450s, fierce competition for the throne between Shiro, son of King Kinpuku, and his uncle, Furi, led to the death of both incumbents and the burning of Shuri Castle. This fire of AD 1453 is of great archaeological significance and is described below. A second fire broke out in AD 1459 at the time of struggles involving the powerful chiefs Gosamaru and Amawari.
The Second Shō Dynasty began in 1469 AD. The reign of Shō Shin (AD 1477 to 1526) was a period of great prosperity for Chūzan. It was Shō Shin who fully established the kingdom’s system of rule (Okamoto 2008, 55). During his reign, all aji were made to reside in Shuri, and their estates were put in the hands of (p.237) stewards (Sakihara 1987, 175). Until that time, most of the aji lived on their own hereditary estates (Kerr 1958, 106), and the change in policy effectively separated them from the rural communities that had been their bases of power. Another effect of coercing the aji to move to the capital was to reinforce the divisions of social class over local kin groups. The Second Shō Dynasty could be seen to be the threshold of true state organization. In this case, the earlier polities of the thirteenth to mid-fifteenth centuries would be considered well-developed chiefdoms.
The Okinawan policy was more radical than that of the Japanese Tokugawa during the eighteenth and nineteenth centuries AD, for Japanese lords were allowed to reside in their own fiefs in alternate years. Weapons were gathered in the capital, and a system of administrative units, magiri and shima, was established (Asato Susumu 2003b). Okamoto is of the opinion that the decline of Ming tributary trade was a factor in the centralization of the Ryukyu Kingdom, in which the state acted as a trader while there was a decline in private trade (Okamoto 2008, 55).
Shō Shin made a concerted effort to develop centralizing cultural institutions, including the hierarchy of state-sanctioned female religious practitioners (noro) who were located throughout the realm (Uezato 2010, 90). From the late fifteenth century the production of luxury goods and ornaments including curved pendants (magatama) for the noro and gold casting as well as lacquer production were centralized under state control. Shō Shin sponsored impressive Buddhist temples such as the Enkakuji as a memorial temple for his ancestors (see below) and erected inscribed monuments around the capital to create a landscape of power (Takara 1987, 84–85). Many concepts of the classical Chinese jinei (Japanese kinai) system that had been adopted by Yamato in the seventh century AD can be found in the Ryukyu Kingdom, particularly during the reign of Shō Shin. These include the designation of three inner magiri (administrative districts) (Nishihara, Mawashi, and Haebaru) to form the heartland (kinai) ssurrounding the Shuri capital (Makishi 2012).
Within the aristocracy residing in Shuri, court status distinctions based on six colors for clothing and accessories were introduced in AD 1506, apparently to rationalize the positions of various outlying local aristocrats into one unambiguous system at the center. Three years later, sumptuary rules were instituted, regulating dress and other manifestations of rank in Shuri (Kerr 1958, 110).
Rebellions occurred in the outlying islands from about the middle of the reign of Shō Shin, perhaps precipitated by harsh exploitation by the Shuri government as it became overextended. The Yaeyama Islands, located some 370 km southwest of Okinawa, stopped sending tribute to Okinawa in AD 1496 or 1497. They were punished in AD 1500 by a force of three thousand soldiers in forty-six (p.238) ships organized by Shuri, who annexed the islands and began to extract tribute shortly thereafter (Sakihara 1987, 185; Pearson 2003; Ōhama 2005).
Shō Shin also reorganized elites into units called hiki, which supervised defense, internal security, and the administration of trade, including the security of official ships that were armed against pirates (Smits 2010; Takara 1987, 102–122). In AD 1546, Shuri Castle was enlarged and strengthened. By this time, the only functioning castles remaining besides Shuri were Nakijin Gusuku, which protected the north under a governor appointed from Shuri, and the defensive castles for Naha Harbor, Yarazamori Gusuku and Mie Gusuku, both built in AD 1554. A further development in 1556 was the creation of a Council of Three (Sanshikan), which developed out of a council of regents. Its members were powerful scholar aristocrats who worked with the king to direct the heads of different government bureaus.
The conquest of Okinawa by Satsuma in AD 1609 brought the end of independence. Although the ruling Shimazu family of the fiefdom of Satsuma held a nominal title, Lord of the Southern Islands, conferred by the Kamakura Bakufu in AD 1206, it had never been able to exercise any de facto control. Ryukyu refused to contribute soldiers to General Hideyoshi’s campaigns in Korea in the 1590s, finally making a delayed contribution of supplies. Following the victory of the Tokugawa Clan in AD 1600, Satsuma repaired its relationship with the first Tokugawa ruler, Tokugawa Ieyasu, who confirmed the hereditary titles of the Shimazu Clan (Kerr 1958, 154–168). In AD 1609 Satsuma invaded Ryukyu and removed King Shō Nei to Japan for three years, extracting an agreement that Satsuma would control Ryukyu and appropriating all islands from Yoronjima to the Amami Islands. Ryukyu was to continue its tributary status with China, giving Satsuma access to China and creating a relation of dual subordination by both China and Japan.
In 1997 I proposed that, in the period of its prosperity and independence (Early Ryukyu Kingdom Period, AD 1429 to 1609), the Ryukyu Kingdom displayed some characteristics of a city-state (Pearson 1997b). Griffeth and Thomas (1981) described the following defining attributes. These states have a defined core, usually enclosed by walls or surrounded by water, and they are economically self-sufficient, often through the acquisition of an immediate and productive hinterland. They are also politically independent and self-governed. They seem to develop when polities in surrounding areas are relatively weak. In the fourteenth to sixteenth centuries AD, central authority in the main islands of Japan was relatively weak, and China had little interest in conquest of such small islands. When power in the Japanese main islands was consolidated and centralized in the early seventeenth century AD under the control of the Tokugawa (p.239) Shōgunate, the Ryukyu Kingdom was subjugated by a Tokugawa vassal, the Shimazu Clan, and lost its independence.
Is it heuristic to think of the Ryukyu Kingdom in terms of a city-state? Comparison of the Ryukyu case with city-states emphasizes its mercantile nature, small scale and hinterland, centralized administration, and opportunistic rise and decline. It was also self-sufficient for food but not for raw materials such as metal. Unlike Mediterranean city-states, it did not have a democratically organized citizen class (polis) or slaves, nor was it ruled by a council of merchants, in the manner of Sakai, Osaka, in the fifteenth and sixteenth centuries AD. In addition, the degree of urbanization in the region of Naha and Shuri in the fifteenth to seventeenth centuries AD was probably low. European and Mediterranean city-states usually occur in competing groups, which is not the case in the East China Sea. However, the Mediterranean city-states are not considered to be the only form of city-state (Charlton and Nichols 1997, 4). In contrast to territorial states, which integrate large areas through a hierarchy of administrative centers, the Ryukyu Kingdom was highly centralized, although there were regional centers for the collection of tribute in Ishigakijima, Miyakojima, Kumejima, and Amami Ōshima, from the sixteenth century AD. Despite its diminutive size, it relied on a relatively large peasant class for agricultural produce (Asato 1998, 137–238). Exploring diversity among small commercial entrepôts in Asia and other areas may lead us to understand the varied processes of their development. Probably the label “city-state” is not heuristic.
The Configuration of the Shuri Capital
In this discussion I refer to the Shuri capital as Shuri Gusuku, to emphasize its Okinawan features, such as type of construction and religious significance, which are distinctive from those of Japanese or European castles. The castle is situated on a ridge of limestone some 100 to 135 m in elevation, lying inland and to the west of the port of Naha. It commands a view over much of the southern part of the island, including the earlier capital of Urasoe to the north and the trading port of Naha to the west, as well as important offlying islands. Its dimensions are 400 m east to west and 270 m north to south, with a total area of 42,000 sq m (Fujimoto and Naka 1980, 288). Beginning with the First Shō Dynasty, it was the castle site of the king of Ryukyu. The oldest inscribed stela in Okinawa, the Ankokuzan Jukaboku no Kihi, was erected by King Shō Hashi in AD 1427 (Shimajiri 1983) to the north of the Sonohyan Utaki near the first gate (Shurei no Mon) of the Shuri Castle. It confirms that the castle was built at least by the early fifteenth century AD. It records the planting of trees and flowers (p.240)
on Ankoku Hill, created with fill from digging the adjacent Dragon Pond (Ryūtan), as well as the political ascendancy of Chūzan and the establishment of tributary relations with the Ming dynasty (Figure 9.1).
The earliest phases of occupation and the initial phases of castle construction are not clear, and recent reconstruction of some of the buildings may make it impossible for archaeologists to expose large areas of earlier strata. The castle itself may date from the reign of Satto (AD 1350–1395), but less complex elite residences probably preceded it. It is likely that Shuri was not the center of Chūzan when it was first built. At that time Urasoe was still the capital. The transfer of the capital from Urasoe to Shuri took place during the reign of King Satto (Sakihara 1987, 105). The finding of very valuable Chinese ceramics, which must have been part of the gifts given by the Ming emperor Hongwu to Satto in AD 1374, in at least two areas of Shuri, means that by the late fourteenth (p.241) century Shuri must have been the capital, not Urasoe (Kin 2008a, 42). Along the outer periphery wall were four outer gates, and nine interior gates gave access to different divisions within the palace. Most of the inner walls were of cut limestone, 6 to 12 m high, surmounted by wooden structures. The arched construction of the gates is of Chinese style. Administrative functions were carried out in three large buildings arranged around an open space (Unā), which was used for state occasions. These are the Seiden (administrative palace), Nanden, and Hokuden. To the east were two residential buildings for the king and queen.
Shuri Gusuku was surrounded by the residences of royal relatives and retainers, temples, and administrative offices (Seto 2010b). To the immediate north was the Enkakuji temple. Outside the eastern edge of Shuri Castle was the Sakiyama Utaki and to the west, the Tenkaiji temple and the Royal Mausoleum (Tama Udun) (Zaidan 2003, 274–277). In the Early Gusuku Period, gusuku were separated from villages, but in the Late Gusuku Period, in the Chūzan area, the gusuku became the nucleus of a walled town. At the time of the First Shō Dynasty (AD 1406 to 1469), additional buildings already occurred inside the walls. In the reign of Shō Shin (AD 1477 to 1526) storehouses for containing the arms that were relinquished by the local aji and residences for the elite were built. To strengthen its fortifications, walls were added in the middle of the fourteenth century AD, during the reign of Shō Hashi. During the reigns of Shō Shin and Shō Sei (reigned AD 1527 to 1555) in the Second Shō Dynasty, the stone walls on the north and east sides were doubled in length during a major enlargement and reorganization (Asato Susumu 1998, 47), achieving a form that remained unchanged until the total destruction of the site in the Battle of Okinawa in AD 1945. Detailed maps of the surrounding area have survived from the nineteenth century, and exploratory excavations are ongoing. An example is the residence of the crown prince, the Nakagusuku Udun, which was built in AD 1621. It has yielded trade ceramics from the late fourteenth and early fifteenth centuries AD. Destroyed in World War II, it became the site of the Okinawa Prefectural Museum from 1966 to 2007, before the opening of the new museum in its new location in Shintoshin, Naha (the former residential area of high-ranking American military officials) (Okinawa Kenritsu Maizō 2010).
In the centuries following its first construction in the early fourteenth century AD, Shuri Gusuku was burned four times, and a portion was rebuilt in AD 1992. The buildings that were standing in early AD 1945 had been rebuilt in AD 1712–1715, and their partial reconstuction in 1992 was based on the eighteenth- to nineteenth-century version. Before the eighteenth century AD it certainly was not exactly the same as later versions. From excavations of the
(p.242) Kyōnouchi, one can see pits dug into the coral limestone and can find Korean-type roof tiles, which appear to be associated with the buildings of the fourteenth to fifteenth centuries AD. Roof tiles of three distinctive types, Korean, Yamato, and Chinese, have been found in Shuri. Although it was previously thought that the so-called Korean-style tiles were made in Korea and shipped to Okinawa, petrographic analysis has shown that the three types were all made with the same clay, which contains coral inclusions and matches clay samples from the western part of Nago City, Okinawa (Yamamoto Masa’aki et al. 2007). These three types, as well as soft gray pottery and floor tiles from Shuri Gusuku, have the same petrographic signature and were all produced in Okinawa. The paste has chert inclusions from the Nago Geological Layer of the Motobu region of northern Okinawa. The Korean-style tiles share some motifs with the Yamato Type, indicating that, after the adoption of Korean manufacturing methods, there was local innovation. Uehara (2004) proposed that either Yamato craftsmen came to Okinawa or Okinawans learned tile production in Japan.
Spatial Organization of Shuri Castle
Different buildings and areas of Shuri Castle served a number of functions in the administration of the kingdom, judging from the prewar plans of the last stage of Shuri, rebuilt in AD 1712–1715 (Figure 9.2).
At least four buildings set around a large forecourt were devoted to record keeping, administration of sacred places, foreign relations, and records of births and deaths. Shuri Gusuku contained the residences of the king and queen as well as the seat of administration (Seiden). The plaza in front of the administrative palace was an area for state functions and assemblies. The administrative palace faced to the west instead of the south as it did in Urasoe (Asato Susumu 2006, 91). This shift in orientation followed Ryukyuan ideology of the king as the child of the sun (tedako), which came to take precedence over Chinese rules of north-south orientation in the early fifteenth century AD (Asato Susumu 2009). On the north and south sides of the plaza were the Hokuden and Nanden, reception halls for Chinese and Japanese delegations respectively. Spaces for ritual activities were dispersed. The largest of these, the Kyōnouchi, was located in the southern portion of the western half of the castle, while the king’s private shrine, the Shinbyōden, was at the extreme east. An important shrine, the Sonohyan Utaki, was located outside of the castle walls on the west, inside the ceremonial gate (Shurei no Mon) and near the juncture of roads leading from the castle to the port of Naha. A watch tower, bell tower, and craft workshop area were situated on the extreme western side. Below the castle walls, on the north side, were the Ryūtan Pond and the largest Buddhist temple in the (p.243)
1. Ceremonial Gate (Shurei no Mon), first erected AD 1529
2. Monument marking beginning of stone paved road to Shikina Summer Residence (Madamaminato Himon), first erected AD 1522
3. Start of Madama Road
4. Monument dedicated to Sixth Abbott, Enkakuji temple, erected AD 1522
5. Shrine (Sonohyan Utaki)
6. Bell Tower
7. Western Parapet
8. Iri no Azana Area
10. Takayosōridun Area
11. Shrine (Madamamori Utaki)
12. Dragon Watercourse
13. Royal Records (Keizuza)
14. Foreign Relations Records, Protocol (Yubutsuza)
15. Royal Storehouse (SK01)
16. Administration of Sacred Places (Jiinza)
17. Registry of Births, Deaths, and so forth (Daiyoza)
18. Warehouse for Ritual Objects (Zenigura)
19. Sundial Clock
20. Shrine (Shurimori Utaki)
21. Storage of Medicine, Tobacco (Naden)
22. Watch Tower
24. Reception Hall for Chinese Envoys (Hokuden)
25. Reception Hall for Satsuma Japanese Envoys (Nanden)
26. Royal Residence (Shoin)
27. Administrative Palace (Seiden)
28. Sitting Rooms for King, Queen
29. Shrine (Kaimei Utaki)
30. Hall for King’s Succession Rituals (Yobokoridun)
31. Queen’s Residence (Ōgon Odon)
32. Royal Refectory (Sakiman?)
33. King’s Residence (Nikaiden)
34. Sitting Rooms for Court Ladies
36. Residence of Queen Mother, Court Ladies (Yosoidun)
37. Kitchen Storage
38. East Parapet
39. King’s Shrine (Shinbyōden)
40. Sashiki Palace (for Queen’s Administrators)
41. Kuninaka Utaki
Excavation areas (approximate scale):
G. Kankai Mon area
H. Outside South Wall
I. Administrative Road Area
(p.244) kingdom, the Enkakuji. When King Shō Shin forced the local lords to relinquish their arms and move to the capital, storehouses for weapons were built inside the castle, presumably in one of the outer enclosures.
The central administrative building, the Seiden, 16 m high in its present form, was the largest structure on Okinawa in its time. By at least AD 1450, it was two stories high. In front of the Seiden were two carved stone dragon pillars erected in AD 1508. In contrast to palaces in Kyoto, which were raised on wooden pillars, the Seiden and other official buildings were set on stone platforms about 90 cm high. They had tile roofs, wooden walls, and round pillars.
Excavations in the 1980s, at the time the Seiden was reconstructed, revealed five different construction periods, but the actual forms of earlier buildings were not clear. Period I, known from only a small excavation sample, was marked by a brown soil layer. Roof tiles of Yamato (Japanese) Type were recovered and are thought to be from a building preceding the Seiden. In Period II, a layer of cut stones was laid down, and its surface shows evidence of burning, perhaps from the disturbance of AD 1453. On top of the cut stones was a building 16.36 m × 9.09 m (Kin et al. 1988, 45–46). The western (front) portion of this building contained burned sherds of trade ceramics, armor slats, metal fittings, and fragments of burned charcoal, suggesting that the structure had suffered extensive burning. In Period III, the stone cutting and paving became extremely rough, as if refacing of the platform was done in a great hurry, perhaps because of some major event that occurred at this time. At least two more refacings occurred, in Periods IV and V, with the facing from Period V remaining until twentieth century AD. The stratigraphy of the base platform of the Seiden has not been tightly correlated to historical events (Zaidan 2003, 276; Ijū 2010).
The Shoin Area
In the area of the guard station (Tsumesho), there was an unusual feature consisting of a pavement of large ridged clamlike shells, hirejako (Tridacna squamosa), placed with the exterior of the shell upward (Okinawa Kenritsu Maizō 2005b, 212). This kind of feature has not been found in elsewhere in Shuri Castle. Beneath this pavement is a layer containing artifacts, the most recent of which is a Ming dynasty Yongle coin dated AD 1408. In the sand and gravel layer covering the pavement there were ceramics dating to the fifteenth to sixteenth centuries AD. About 30 cm above the shell pavement layer was stone pavement no. 2, which resembles the paved walkways in other parts of the castle. It would (p.245) be difficult to walk on the shell pavement; therefore its meaning and purpose are not clear.
The Nanden (Southern Palace) and Hokuden (Northern Palace)
The Nanden and Hokuden flanked the Seiden, at right angles, forming two sides of the main plaza. The fourth side of the plaza was formed by the Hōshin Gate. In the period from AD 1609 to 1879, the Hokuden housed visiting Chinese envoys, while the Nanden housed visiting Satsuma envoys, so that the configuration reflected the dual sovereignty of China and Satsuma in relation to the Ryukyu Kingdom. The Hokuden was constructed in Chinese Ming style. Associated artifacts and stratigraphy confirm that the configuration of the three buildings was in existence from the mid-fifteenth century AD.
Six building periods of the Nanden were discerned from excavation (Okinawa Ken 1995b, 173). Nanden Period VI appears to be the building of the Satsuma Period, after AD 1609. The excavation reporters mention that there is a reference in the Ryūkyū koku yuraiki (Hokama and Hateruma 1997) to a building called the Gusuku Udun, which may be one or all of Nanden I to VI. Nanden Period I is correlated with Seiden Period II. It was built on a leveled outcrop of Ryukyu limestone. The east corner of the kidan platform remains. It was constructed of carefully cut limestone blocks, set in the nunozumi style. The surface of the corner has been reddened by strong heat, which is thought to be from the fire of AD 1453. Nothing above the stone base remains. The remains of Nanden Periods II, III, IV, and V appear to be contemporary with Seiden Periods III, IV, V, and VI.
The building sequences of the Hokuden are not clear, but there seems to have been a building on the site at the time of Seiden Period II. Artifacts recovered from this area confirm that the three buildings were in place by the mid-fifteenth century or slightly earlier. Although some stone adzes found at the site of the Hokuden may be specimens belonging to a museum housed in the Hokuden before World War II, most artifacts recovered are from the fourteenth to fifteenth centuries AD, Chinese celadon comprising 50 percent of the materials. East of the Nanden was the Ouchibaru Shoin, a residence of the king and queen that was connected to the Seiden. Behind this was the Shoin, a Japanese-style reception building with a tiled roof.
The Royal Residential Area
In an area to the immediate south of the Seiden, in the space between the Seiden and the Nikaiden, were several small buildings that were the living quarters of (p.246) the royal family. There are no existing photos or historical records of these buildings. Some stone alignments and pillar bases have been found as well as artifacts from the fourteenth to nineteenth centuries AD including ceramics, roof and floor tile fragments, metal fittings, and beads (Okinawa Kenritsu Maizō 2007b). The first building, the Yoseman, a kitchen and bathing area for the king and queen, could not be conclusively located. The second building, the Ōgon Odon (Yellow Gold Palace), contained the sleeping quarters for the king and queen. Although the date of construction is unknown, there are records that it burned in AD 1660 and 1709 (ibid., 206–210). No features can be conclusively dated to the period before AD 1609. It is thought to have had a tile roof similar to that of the Nikaiden. A substantial number of artifacts, including roof tiles, floor tiles, Chinese celadons, brownwares, and Okinawan ceramics as well as Thai, Vietnamese, Bizen, and other Japanese main islands wares were found in this area. In particular, sherds of large Chinese vessels were recovered, dating to the fourteenth to nineteenth centuries AD, with the majority dating to the fifteenth century AD. The remains of a third building, a kind of side room for the ruler’s personal assistants such as administrators and scribes, the Kinju Tsumesho, is located to the southwest of the Yellow Gold Palace (Ōgon Odon). Its paved floor and pillar bases were located in the excavations. In the space between the Kinju Tsumesho and the Ōgon Odon there was a room called the Suzuhiki no Ma (lit. Bell-Pulling Room) where female officials could summon attendants to the ruling family. As with other buildings in this area, there are no detailed written records.
The Nikaiden (Main Royal Residence)
The Nikaiden, now reconstructed, is located on the south corner of the Seiden. Historical records date its present form to the eighteenth century AD (Okinawa Kenritsu Maizō 2005e, 93). On the west a staircase connected it to the Seiden, and there was also an eastern staircase that gave access to the outside.
Although historical records pertaining to this area of the castle do not extend back beyond the eighteenth century AD, excavations have yielded a stratigraphic record extending back to the late thirteenth century AD. Layer I contained postwar debris, mixed with disturbed material that included ceramics of the Gusuku Period. Layer II belongs to the construction period of the Nikaiden and includes its stone base. Below the stone base is an unconsolidated deposit of sherds. It contains burned ceramics, particularly celadon sherds, which date to the fourteenth and fifteenth centuries AD. This dumped deposit, consisting of ceramics, coins, metal objects, and beads, appears to have been transported and dumped as fill (ochikomi). Thus it is not clear that the contents were originally (p.247) associated with a royal residence, but it seems likely that they were not transported very far. The contents are described below.
Nikaiden Dumped Deposit: Archaeological Features Sr11, Sr12, Sr24, Sr32, Sr33
In the Nikaiden Dumped Deposit a group of ceramics comprising blue and white, celadon, and whiteware is particularly notable. The celadons and white-wares are comparable to those found in the Sinan sunken ship found near Mokpo, Korea, dating to the 1320s AD. The Chinese blue and white wares fall into two groups, dated to the Yuan and Ming periods. The Yuan blue and white sherds are particularly large and come from large specimens. The total sherd count is 18, and the minimum number of vessels is 10, while the total count of Ming sherds was 370 and the minimum number of vessels was 71 (Okinawa Kenritsu Maizō 2005e, 55). Some of the Yuan blue and white finds from the Nikaiden are unique in Okinawa Prefecture. These include blue and white jars (tsubo) with a lotus pond motif as well as a tree peony and arabesque motifs, a bottle (usually referred to by the Chinese yuhucun) with dragon and pearl motif, and a small rectangular stand with open work. Also unique is a whiteware bowl of the Chinese Shufu Type. How did these particularly unusual ceramics reach Shuri? Were they actually traded during the Yuan period, or were they tributary items that were already antiques when they reached Okinawa? Kamei (2008; 2009) emphasizes that a substantial number of Zhizheng Type finely painted specimens, which are different from crudely painted small containers found in Southeast Asia, have been found in Shuri and that their quantity is comparable to the numbers found in Southwest Asia and China. Were the Yuan blue and white ceramics of the Nikaiden part of a separate exchange system in which Shuri was a major player? Shuri Gusuku has yielded 79 Yuan blue and white vessels, 72 percent of a total of 109 from seven sites in Okinawa. Of the 459 total specimens of Yuan blue and white vessels recorded in a comprehensive survey by Kamei from China, Korea, Japan, Ryukyu, and South and West Asia, 109 come from Ryukyu and 179 from South and West Asia (Kamei 2009, 66). Kamei believes this distribution implies that the Ryukyus were an important center for the distribution of these ceramics, which have often been considered diplomatic gifts produced for distribution in South and West Asia. The pieces are particularly large and impressive. Kin (2008a) has suggested that they may have been part of the initial tributary gift from the Ming rulers. The large finely painted pieces are similar in type and quantity to those found in the Majapahit capital of Trowulan, Java (Dupoizat and Harkantiningsik 2007, 57–64)
Two small brown-glazed jars of a type not often found in other parts of Shuri (p.248) or Okinawa Prefecture have been identified as products of the Cizao kiln, Quanzhou, Fujian. This type has also been found in the Hakata Site Group and in Dazaifu, and is dated to the twelfth to thirteenth centuries AD. Early connections with Quanzhou are particularly interesting in light of the discoveries from Urasoe (described in Chapter 7). Two other wares are thought to be products of kilns in the lower Min River area of Fujian: a brown-glazed jar and a white-glazed jar (Okinawa Kenritsu Maizō 2005e, 96).
The deposit contained 532 base sherds of bowls, with black oil spot glaze termed tenmoku in Japanese, on a light gray body and a sharply trimmed foot rim, made in Nanping City and other kilns in the area of Fuzhou, Fujian Province. They are of the most common type found in Okinawa Prefecture. There are also sherds of tea containers (caddies). They are dated to the fourteenth century, estimated to be from a relatively short period. Could they represent a single tribute gift, shipped from Fujian? These bowls confirm the consumption of tea, in Chinese style, by the Ryukyu court. Alternatively they could have been reserved for shipment for Japan (Seto 2004, 146).
Vietnamese wares dating to the fourteenth and fifteenth centuries AD were far more numerous in this site than in other Okinawan or Japanese sites. A total of 545 sherds or a minimum number of 105 pieces was recovered, comprising 4 percent of the total sherds from the Nikaiden. Blue and white pieces, such as bowls with an unglazed base with chocolate-colored interior ring, were particularly numerous. The bowls are of two types, with different dating (Okinawa Kenritsu Maizō 2005e, 94). Type I is a deep bowl with everted lip. The white glaze is close in tone to that of Yuan blue and white. By comparison with specimens from the Dazaifu Site in northern Kyushu, it is dated to the mid-fourteenth century AD. Type II is dated to the late fourteenth to mid-fifteenth centuries AD. The Thai wares, which also comprised about 4 percent, were primarily sherds from large brownware jars, but there were a few sherds from earthenware jars with Chinese characters stamped on the shoulder, which may have been produced in Laos in the fifteenth century AD (Okinawa Kenritsu Maizō 2005e, 96).
Finally, there are a few sherds of soft grayware unfired braziers (late fourteenth to early fifteenth century AD), which appear to have been made in a site on the Inland Sea coast of Kyushu. Similar wares have been found in Ichijōdani and Tosaminato (Okinawa Kenritsu Maizō 2005e, 96).
The ceramic deposits of the Nikaiden define an archaeological space that would otherwise provide few clues as to its function or program. The Yuan blue and white ceramics indicate the privileged role of the Ryukyu kings in the Chinese tributary system and also show their power to conduct exchange with peers in East and Southeast Asia. The Fujian tea bowls indicate the adoption of tea (p.249) drinking that they shared with other elites in East Asia. In addition to tenmoku tea bowls, tea containers (cha tsubo) are found not only in gusuku sites but also in large villages (Arakaki 2004; 2007). Tea drinking is not only closely associated with China and Buddhism, but is also linked to mediaeval Japanese samurai culture. The ceramics from Thailand and Vietnam support the accounts of elite exchange in the records of the Rekidai hōan (Okinawa Kenritsu Toshokan 1992).
In a layer termed SB Lower Layer, found under a stone pavement in the Nikaiden area, another accumulation of ceramics yielded a minimum number of seventy-seven vessels, mostly celadon, with a few tenmoku and whitewares, dating from the fourteenth to fifteenth centuries AD. The celadons are largely plain with everted rims, and others have broad, unridged lotus petal motifs on the exterior, as well as “lama-style” lotus petals. Six tenmoku bowls were stuck together with a calcareous deposit from limestone; they may have been redeposited from another area of the castle site. Many of the ceramics appear to have been exposed to strong heating. They may have gone through the same fire as the ochikomi material and the Kyōnouchi artifacts (described below), and may actually belong to a shorter time span than the others.
The Kyōnouchi was a ritual space of the mid-fifteenth century AD (Okinawa Ken 1998; Okinawa Kenritsu Maizō 2009b). Its name means “sacred area with spiritual power” (Okinawa Ken 1998, 15). Archival evidence suggests that there were originally five utaki (shrines) in the area. Of the total area of 5,000 sq m, 2,000 sq m have been excavated. A substantial number of features, including walls, alignments, pavements, ditches, pits, building foundations, and a staircase have been excavated. From the complex configuration of features, six periods of construction were determined (Okinawa Ken 1998, 26) (see Figure 9.2). An investigation of the features from this oldest part of Shuri Castle yielded earlier features dating to the first half of the fourteenth century AD, including stone building material, pillar bases, and drainage ditches. Roof tiles of the same types as those found in Urasoe have been found in this area. The Kyōnouchi was referred to as the old gusuku (Kogusuku). Tawada has proposed that the oldest buildings were built in the Kyōnouchi area at the time of the move of the capital from Urasoe to Shuri in the fourteenth century AD (Tawada 1980; Kinjō Kamenobu 2001, 72–73). The area where these early structures are located is termed (p.250) the Takayosōridun. In that area, the remains of a multistoried building thought to be the Takayosōridun itself was found in excavations, but it was heavily disturbed. According to Kinjō Kamenobu (2001), after the Takayosōridun burned, the Kyōnouchi assumed a religious function. It was separated from the structures with primary political functions such as the Main Palace, South Palace, North Palace, and Hōshin Gate by stone walls 3 to 7 m in height. Located at one of the highest points within the castle, it had a lookout from which Kudakajima and Kumejima as well as Zakimi on Okinawa Island could be seen.
Early Features (Sa24, Sa03)
On the south edge of the Kyōnouchi, there are low walls and pits that contain sherds of Chinese whitewares and tenmoku dating to the early and mid-fourteenth century AD (Okinawa Ken 1998, 29). These indicate that there were substantial buildings within Shuri Gusuku before the unification by Chūzan and the establishment of the Ryukyu Kingdom. These finds confirm Tawada’s hypothesis mentioned above.
A cave feature had cut stones at the entrance and a floor of sandstone overlain by gray tiles. The main use of the cave appears to have been in the sixteenth and seventeenth centuries AD. In the lowest layer of the cave floor, a sherd of Vietnamese blue and white porcelain was found (Okinawa Ken 1998, 25–26).
Features B13, B14, C15, C16, D15, and D16 were six fourteenth century AD pits, with a depth of about 30 to 40 cm and a diameter of 40 to 50 cm, that contained cow crania and hooves thought to be the remains of some kind of ritual (Okinawa Ken 1998, 27). A similar group of pits has been recovered at the Nakamamura Site, Haebaru Chō, Okinawajima, a thirteenth century AD site yielding the lower mandibles of five cows lined up next to the remains of a kuba tree (Livistona chinensis). These are also thought to be the remains of an agricultural ritual (Kinjō Kamenobu 2001, 78). The kuba, or Chinese fan palm, is still believed by Okinawans to be a resting place for spirits, and groves of this tree often surround shrines (utaki), religious sites sometimes sheltered in a small structure, marked by an incense burner and small porcelain bottles for flowers or water. Further ethnographic research is required to determine the significance of this ritual and its location within the Kyōnouchi.
The most important feature excavated in the Kyōnouchi is the remains of the Royal Storehouse (sōko), thought to have been burned in AD 1459. A portion of the storehouse consisting of a rectangular pit or basement 3 m × 4 m and 0.5 m deep, with three steps, shows signs of burning. Nonceramic artifacts include armor, a sword, bead ornaments, and a metal lock (Okinawa Ken 1998, 255–258). The Omoro sōshi songs (Hokama and Saigo 1972) mention the ritual significance of the sword and armor. The beads may have decorated bottles of a type seen in later heirlooms.
The major contents of the storehouse are ceramics, including some very fine specimens (see frontispiece). The contents of the storehouse are thought to have come from two main sources. First, an important collection was received from the Chinese court in AD 1459, and, second, ceramics were received in trade with various countries in the mid-fifteenth century AD. A fire destroyed the Royal Storehouse in AD 1453 at the time of the Shiro-Furi Rebellion, a succession struggle between Shiro, the son of King Kinpuku, and his uncle Furi, the sixth son of the previous ruler, Shō Hashi. To replace the ceramics stored there, King Shō Taikyu received a new collection from the Ming court, in exchange for copper coins, rather than the usual gift of sappanwood from Southeast Asia. Sixty percent, by value, of the goods from China came from the Peking Treasury, and 40 percent came from Fujian in the form of a special kind of woven silk and locally made celadon and blue and white ceramics. Since the 6:4 division took account of the value of the objects, the actual number of pieces from Peking was relatively small (Kinjō Kamenobu 2001, 78; Okinawa Ken 1998, 256). Thus, the collection contains some dramatic treasures of the highest quality along with large numbers of ordinary objects such as celadon bowls (Okinawa Kenritsu Maizō 2001c). In the frontispiece, celadon vessels are grouped on the right. They consist of jars, ewer, stemmed cup, dishes, and a large baluster vase (rear), dating to the early fifteenth century AD. On the left are blue and white vessels. The vases and large jar (rear) are dated to the fifteenth century AD, whereas the large covered box and stemmed cup are dated to the fourteenth century AD (end of the Yuan dynasty). The overglaze enamel bowl in the center foreground is dated to the early fifteenth century, and the red-glazed ewer (center) is dated to the fourteenth century.
Spectacular pieces of the royal collection include a whiteware bowl referred to as the Shufu Type, a Ding Ware white lotus shaped cup, a red-glazed Yuan ewer, a ruri-(brilliant blue) glazed ewer, a dragon pattern blue and white stirrup cup (bajōhai), an exceptionally large Yuan blue and white covered box (Kamei 2002a, b, and a celadon bottle with tree peony and arabesque (Japanese karakusa (p.252) ) decoration. Examples of several of these are previously unknown in Japan. The red-glazed ewer is thought to be unique worldwide; and the white stirrup cup with stamped dragon pattern, thought to have been presented by the Ming court to legitimize the authority of surrounding vassal states, is one of twenty specimens known worldwide. Many of the Yuan and early Ming celadons are of exceptional quality.
In the period between the rebuilding of the warehouse after the 1453 fire and the second fire in AD 1459, ceramics were received from Siam, Vietnam, and Japan, and this mixed collection was burned in the second fire. Thai earthenware, in the form of sixty-seven lids and four vessel sherds, was recovered. The lids are thought to have been used for large brown-glazed four-eared jars. Some may be of Okinawan manufacture (Okinawa Ken 1998, 61). Kin’s explanation is that large brown- and black-glazed Thai jars with earthenware lids contained special Siamese liquor served to Chinese envoys (Kin 1993; 1999; 2000). Mukai (2002, 79) suggests that they may also have been used as containers for other products from different parts of Southeast Asia such as food or dyes. Kin (2000) proposed that, since most of the Kyōnouchi ceramics are food-serving vessels of Chinese manufacture, they must have been used for serving Chinese envoys who arrived in the tribute missions or in palace rituals such as succession ceremonies. Kinjō Kamenobu proposed that some of the pouring vessels could have been used in religious ceremonies associated with local shrines (utaki) (2001, 78). Kin (2008a) notes that the finest ceramic pieces in both the Kyōnouchi Royal Storehouse and the Nikaiden Deposit cover a time span from the mid-fourteenth to mid-fifteenth century AD. He proposes that the late Yuan pieces were part of the gift of ceramics given to King Satto in AD 1374 by the Ming emperor Hongwu. In both interpretations the finest wares were presented by the Ming court.
Fragments of royal armor, including helmet ornaments, were found in the Royal Storehouse. These are important for showing the construction of a distinctive Ryukyu style in the first half of the fifteenth century AD. Yamamoto and Uezato (2004) state that, although many finds of metal armor are yet to be systematically studied, current finds show that Ryukyu armor is generally of Japanese style. The general form of the armor follows that of the Japanese main islands, but some of the decoration is distinctive. The helmet ornament (tatemono) (Figure 9.3) has a main column, and two hoe-shaped wings. The shape, unique to Ryukyu, consists of sun, moon, and stars. This motif is termed the motif of auspicious clouds, moon, sun, and stars (zuiun hitsuki sei mon). Fragments of similar helmet ornaments have been found in the Nanden and the Hokuden. The metal helmet in Japanese armor is not a single cap but is composed of metal-like strips fastened together. Examples from Shuri show a distinctive Ryukyu feature in which the helmet was composed of an underlying set (p.253)
1. Deer-antler-shaped uprights (The Japanese term describes them as hoe-shaped, kuwagata). 2. Main upright decoration (tatemono) with Ryukyu motif of clouds, sun, and moon. 3. Top of helmet (tehen) with hole where metal slats meet. 4. Chrysanthemum-shaped metal fitting covering hole. 5. Tassel held in riveted ring. 6. Decorative band in shape of sacred boundary fence (ikaki). 7. Helmet (hachi) composed of metal strips. 8. Skirt of lamellar pieces for protection of neck. 9. Fastening cord. 10. Ornament with sawtoothlike edge (kyomon kanamono). 11. Visor (mabisashi). 12. Ornamental socket for upright ornament. 13. Protective flange (fukikaeshi).
of iron strips overlain by a layer of bronze strips, which were probably gilded. The creation of unique forms of armor of Japanese style confirmed the status of the king of Ryukyu as a paramount among Japanese equals at official functions with Okinawan confederates or Japanese envoys, in a similar manner to the use of Chinese-style royal regalia of appropriate rank in ceremonies for the Ming envoys. The unique symbolism of the helmet ornament reflects Okinawan political identity and cosmology.
The Shurimori Utaki was situated to the west of the other utaki in its own walled enclosure entered through an arched gate (Kinjō Kamenobu 2001, 74, 79). It was devoted to the worship of the legendary founding people of Okinawa, the Amamikyo, and was one of the most sacred shrines within the castle. Its general location is known from eighteenth-century maps. Excavations were undertaken in hopes of finding remains that would be useful for its reconstruction; however, nothing could be found except a portion of a wall from the rebuilding of AD 1712. While the utaki is mentioned in the early sacred song collection Omoro sōshi (AD 1531–1623) (Hokama and Saigo 1972) and the first Okinawan history, the Chūzan seikan, written about AD 1650 (Haneji 1983), and appears on early-nineteenth-century maps as a C-shaped enclosure (Okinawa Kenritsu Maizō 2008b, 107–108), it disappeared from maps between AD 1881 and 1893. Its complete disappearance “is thought to have some relation to the dissolution of the monarchy” (ibid., 108), suggesting to me that it may have been intentionally destroyed to extinguish Okinawan state religion. Most of the finds, which consist of sherds of ceramics and roof tiles, and Chinese, Thai, and Vietnamese ceramics, came from disturbed layers. It now seems likely that some of the Yuan types of wares were made in the first decades of the Ming dynasty (Lam 2012).
Ritual Ceramics from the Kyōnouchi
In AD 1936 Itō Chuta and Kamakura Yoshitarō made small excavations in two locations on the northwestern side of the Kyōnouchi (Itō and Kamakura 1937, 43–48; Kamei 1986b, 368). They found several hundred sherds of water droppers, jars, and other utensils of South Chinese Three-Color Ware, which comprised half of the entire sample of two test pits, suggesting a high concentration of this ware. In recent excavations in the area thought to be the location of the Shurimori Utaki, a total of thirty-nine sherds of the same ware were recovered. The concentration of this ware in the ritual area of the Kyōnouchi rather than in administrative buildings confirms its importance in state religion and religious networks in the fifteenth to sixteenth centuries AD (Kamei 1986b, 368). These wares have been found in other Okinawan sites such as Nakijin and the northern island of Akusekijima; Kamei proposes that since the center of distribution of these wares is Shuri in Okinawa, these objects had an elite ritual function in the Ryukyus and were exchanged in Ryukyu official voyages.
A product of kilns in the Quanzhou region of Fujian Province, the ware was decorated with unusual combinations of lead glazes, such as brown, yellow, and green, as well as other colors (see Appendix 1). Many of the objects are small (p.255) water droppers, used for diluting ink in painting or calligraphy. The mold-made forms take eccentric shapes, such as those of birds, crayfish, small boxes, and plates. Small spouted pouring vessels are also found (Kamei 1986b, 355–374). From the context of these objects in Okinawan sites and sites on the main islands of Japan, Kamei dates them to the fifteenth and sixteenth centuries AD. Although the actual kiln sites producing water droppers are not known, covered boxes with similar glaze, termed Cochin glaze, from excavations of the Hangdian kiln, near Pinghe, Zhangzhou, Fujian, are dated to the later sixteenth and early seventeenth centuries AD (Chado Shiryōkan 1998).
The Iri no Azana Metalworking Area of the Kyōnouchi
In the extreme western area of Shuri Castle, abundant evidence of metalworking dating to the fifteenth to sixteenth centuries AD has been found (Okinawa Kenritsu Maizō 2009b; Uehara 2009a, 40–49; 2009b). The finds include the remains of a furnace for extracting metal that was connected to a shallow trench, crucibles, tuyeres (bellows valves), large amounts of iron and bronze filings, molds for the production of temple bells, and other metal objects. The finding of the stem and branch of a multiple-coin mold, but not the actual coin mold, raises the possibility that three types of coins bearing characters of Okinawan reigns could have been manufactured here. The stem and branch are of the same type of mold as those found in mediaeval sites in the Japanese main islands. Uehara notes that copper slag with the same composition as the material in Ryukyu coins has been recovered. These types of coins, Taisei Tsūhō, Sekō Tsūhō, and Kinen Sehō, have been attributed to the Ryukyu Kingdom around AD 1450, on slim evidence, since they have cast characters that match with posthumous names of three Okinawan rulers, Shō Taikyū (AD 1454 to 1460), Shō Toku (AD 1461 to 1469), and Shō En (Kanemaro) (AD 1470 to 1476). A fourth type, Chūzan Tsūhō, is said to have been minted in the reign of King Satto (Miyagi 2009). Of the 3.5 million mediaeval coins found in all of Japan, Miyagi reports that only 210 specimens of these Ryukyu coins have been reported. They appear to have been made by reworking an old Chinese coin (Eiraku Tsūhō) of the Ming Yongle era (AD 1403–1424) to make a master or by making a new master from which a mold could be produced (Smits 2008; Uehara 2009a). Jones (2007, 85–94) concluded that they were produced somewhere in Kyushu as counterfeit. However, without the actual mold of the coin such casting cannot be proven. Like mediaeval Japan, Ryukyu did not mint its own official inscribed currency until after the seventeenth century AD, with a possible exception of the four types mentioned above. Until the recently reported excavations of the Iri no Azana, no evidence for the production of these four coin types had been recovered from the Ryukyus. If local production can be (p.256) confirmed, it would appear that manufacture of an independent currency for external use was meant to confirm and legitimize Ryukyu as an independent state. Further research on this topic will add a new dimension to the understanding of the political economy of the kingdom and the use of the extreme western portion of the castle.
Other cast objects include a temple gong (unpan) and metal sake bottle. It is thought that the furnace technology for casting these objects came from mediaeval Japan, although there are some similarities with Song Chinese furnaces. From the finding of the outer portions of molds with raised nipples and horizontal bands, it can be concluded that large temple bells were cast in the Iri no Azana under royal control. Crucibles have been found in other parts of Shuri Castle including the Shurimori Utaki (Uehara 2009a, 49). A separate report (Uehara 2009b) describes in greater detail the evidence for metal casting in the Kyōnouchi. Some 1,670 metal fragments were found, including tools such as handles, ornaments such as decorative pieces, ritual items such as small jingle bells, mirrors, incense burners, a long-handled dipper, vases, and utilitarian items such as cauldrons. Many glass beads of various types were also noted. The small bells may have been used along with mirrors and incense burners, since they have been found together in the Shigemajō Enclosure of Nakijin and in the Tenkaiji Temple Site (Okinawa Kenritsu Maizō 2009a, 253). A total of twenty-two Chinese coins dating from AD 1078 to 1205, and AD 1368 to 1408 were also found.
Armor and related ornamental fittings were cast in the Iri no Azana area. These consist of pieces of perforated armor platelets, fragments of chain mail, and cast bronze fittings and ornaments for helmets, which have been found in several localities of Shuri Gusuku, including the distinctive pieces from the Royal Storehouse in the Kyōnouchi area.
The casting of iron and nonferrous metals shows separate patterns of development in Okinawa (Uehara 2009a, b). Ironworking began in the latter half of the twelfth century AD at the latest. Ironworking sites are not in gusuku but in village sites; 150 sites are recorded. Weapons, armor, agricultural tools, fishing tools, building tools, knives, and iron cauldrons were produced. Iron objects that could not be produced locally were imported in the same manner as foreign ceramics. The working of nonferrous metals began in the thirteenth or early fourteenth century AD. Workshops for nonferrous metals also worked iron, making religious objects as well as ornaments for armor. Many of the copper or bronze objects are ritual objects the casting of which was under the tight control of royal castles (Urasoe and Shuri). Nonferrous metal production can be divided into two periods, thirteenth to fourteenth centuries and fifteenth to sixteenth centuries AD. In the former period it was small in scale and limited to very large castles; in the (p.257) latter, although there are no historical documents, it is clear that bells were cast. More specialized production began in the seventeenth century AD.
The casting of bells for the Shuri Gusuku and surrounding Buddhist temples involved the recruitment of specialists from the main islands of Japan and the marshaling of knowledge and resources. Some scholars have proposed that this kind of metallurgical enterprise could not have been accomplished on Okinawa, and the bells were imported. However, the quantity of bells cast around AD 1450 seems to be too great for them to have been imported, and the finding of fragments of bell molds confirms their local manufacture. Control by the king of the casting of temple bells, a symbolic hallmark of Buddhist temples, may indicate the king’s control of the production of Buddhist spaces. The fact that the bells were produced in an old, sacred area of Shuri shows the power of local religion over Buddhist institutions in Okinawa. At the same time, it seems puzzling that an activity as dangerous as casting bells would have been undertaken inside the confines of a royal capital, with the danger of spreading fire and toxic fumes. Could some of the bells have been brought from the Ōsaka region, on the ships of Sakai merchants?
Along the steep southern face of Shuri Gusuku below the Kyōnouchi area, excavations in 2001 and 2002 yielded a number of old tombs, stone stairs, alignments, and pavements (Okinawa Kenritsu Maizō 2004). Three substantial old tombs were recovered. No. 2 resembles the cliffside mausoleum of Urasoe Yōdore in its side location and constructed access. Judging from its location, it must be an official tomb; however, rulers of fourteenth- and fifteenth-century Shuri were not buried here but were buried in separate locations within the Shuri region. There is also a cave that appears to connect to the Kyōnouchi. This feature is known as Kundagusuku and is thought to date to the fifteenth and sixteenth centuries, since a bronze mirror recovered from it dates to that time period.
The Area Between the Kankai and Kyūkei Gates
The area between the Kankai and Kyūkei Gates is thought originally to have been outside the castle walls but to have become part of the castle after the expansion under Kings Shō Sei and Shō Shin in the latter half of the fifteenth century to the first half of the sixteenth century AD (Okinawa Ken 1988, 139). However, the excavators reported that no fifteenth to sixteenth century artifacts that would confirm this dating were found; instead fourteenth century AD trade ceramics were recovered, suggesting that the dating of this construction should be reconsidered. Evidence of bombing in World War II was found in several places.
(p.258) A tank and conduits from the spring that gives rise to the Kansuigawa (River of Water of the Cold Season) were found. It flowed through a dragonhead-shaped pipe similar to the one in front of the Zuisen Gate, into a tank, and on to the two ponds on the north edge of Shuri Castle, the Enkan (Circular Mirror) and Ryūtan (Dragon Pond). Stone drains around the Zuisen Gate, on one side of the expanded enclosure, were found to be made of volcanic tuff, which is not found locally and must have been brought from Kagoshima (Okinawa Ken 1988, 145–147). Several stelae were erected in this area and in front of the Zuisen Gate. Some fragments of these have been found. Ceramics found in this area include blue and white, whiteware, Koryo celadon, and Yamato Type roof tiles and Meiji Period artifacts including Okinawan ceramics. Gaming pieces or counters, made by shaping broken ceramics into round disks, were found in this area as well as other areas of the castle site. Their diameter ranges from 3 cm to 7 cm, and they were made of roof tile fragments as well as old ceramic sherds. From their wide distribution they are thought to have been used by both adults and children. They were not found in association with any particular features.
Kobiki Mon Area
A small number of hajiki (red earthenware) shallow dishes that look like oil lamps from southern Kagoshima, thought to have a ritual function, were found in the Kobiki Mon area. Their temper may be of local origin, raising the possibility that they are locally made, following a prototype from Kagoshima. They look like a miniature frying pan with a short handle (Seto 2005).
Excavations Outside the Walls, Southern Side
In preparation for an access road extending into the castle from the southeast side, an area of sloping terrain, in which mixed cultural material from the castle has accumulated in depressions, was excavated (Okinawa Kenritsu Maizō 2001b, 208). A wide variety of Chinese, Thai, Vietnamese, Okinawan, soft gray, and Japanese main islands ceramics were recovered. The most common group of Chinese wares was Longquan celadon dating from the fifteenth to seventeenth centuries AD. Other Chinese ceramics included white, blue and white, glazed red enamel, black (tenmoku) wares, as well as South Chinese Three-Color Wares from the Quanzhou area of southern Fujian. The most common forms of this ware were a greenware wine bottle (shukai tsubo), incense burners, and wide-mouth jars. Seventeenth-century Japanese wares included Nabeshima, Hasami, Bizen Type, Seto, Minō, Kyōyaki, Shigaraki, and Satsuma. These wares as well as later (p.259) ceramics from south China and Okinawa show the continuing vitality of the castle after the Satsuma Invasion, a period that is beyond the scope of this book.
The Madama Road
In AD 1552 a military road called the Madama Road was built from the Shurei Mon to the Yarazamori Castle in Naha Port, a distance of 10 km. The road complements the earlier road system running from Shuri Castle downward through Asato to the north side of the Naha port. It was important for the defense of the port against Wakō pirates and for the conveyance of tributary goods to Shuri from the port. At the point of departure from Shuri Castle, on the southeast side of Shurei Mon, a monument indicating the military importance of the road and describing its route was erected. A socket for the base of this monument was cut into the limestone bedrock. Some small fragments of the original monument, severely damaged in the war, were found in excavations (Okinawa Kenritsu Maizō 2005b; 2007a, 69, 77; 2008a). The immediate area surrounding the monument was paved, and, in the disturbed soil, Chinese, Japanese, and Southeast Asian ceramics, mostly postdating the erection of the monument, have been found (Okinawa Kenritsu Maizō 2008a). A sherd of a fourteenth century AD Longquan mallet-shaped (kinuta) vase with fish/dragon handles was noted. This ware is considered to be very valuable by modern connoisseurs, and it is similar to specimens from the Sinan ship dated to around AD 1323 (Information Section 1984). Its relation to the actual monument is not clear (Okinawa Kenritsu Maizō 2008a, 24). The rarity of this type in Shuri suggests that it was deposited intentionally. On either side of the Madama Road were other important stelae.
Between the Shurei Mon and the Kankai Mon there was a long, wide paved roadway and a plaza (hiroba) that must have functioned as an assembly area (Okinawa Kenritsu Maizō 2007a).
In excavations outside the eastern end of the castle wall, archaeologists undertook exploration of the Kuninakagusuku Utaki, which is said to have had royal patronage (Okinawa Kenritsu Maizō 2005c). This area was damaged not only by the war, but also by postwar construction. An utaki exists in this area at present, and it was not possible to dig under it. Trenching uncovered walls and a pavement that may be part of a path leading from the castle. Chinese ceramics from the fourteenth to nineteenth centuries were recovered.
The Tenzan Ryō
that the sarcophagus was hidden from view, it was destroyed in World War II, but drawings remain. At present only the imposing base for a sarcophagus (Figure 9.4) remains in a secluded site in Shuri (Asato Shijun and Morimoto 1984). In an earlier stage of royal burial (thirteenth century AD), in the Yōdore mausoleum at Urasoe described in Chapter 7, the royal burials were displayed in open manmade caves; however, in the renovations of the Urasoe Yōdore Mausoleum in the early fifteenth century AD, the artificial caves were walled up and protected from view by anterior walled courtyards (Asato Susumu 2009). The Tenzan Ryō corresponds to the second stage of protection, thought to reflect the rise of the sun god cult in which the king took on added power, and royal burials were shielded from view. In the early fifteenth century AD both mausolea were furnished with monumental stone sarcophagi imported from Quanzhou, China. The carved base has decorative motifs including turtles and deer, which may be related to Daoist ritual (Tomiyama 1992, 144).
Adjacent Buddhist Temples
Although it seems that Buddhism never became widely accepted by the common people of the Ryukyus, it was important at the official level. China Teikan (2008) has proposed that the Ryukyu Kingdom had a deep relationship with Buddhism; however, since much of the discussion of Okinawan popular religion has been undertaken by students of popular religion and folklore who emphasized local religious practice, the strong Buddhist connection has not been appreciated. The earliest temple, the Gokurakuji at Urasoe, may well have been (p.261) founded by Chinese priests (China 2008, 33–35). Thereafter strong connections were maintained with Japan. A substantial number of temples were built in the late fourteenth and early fifteenth centuries AD, some seventeen being built by the reign of Shō Shin (AD 1477 to 1526). From tabulations in the Naha rekishi chizu (Historical atlas of Naha City) (Naha Shi 1986, 208–212), some forty temples and branch temples are presumed to have been built before AD 1700, mostly in the Shuri and Kume areas. Shuri had at least three Japanese Zen Buddhist temples, all of the Rinzai sect, including the Tenkaiji, the Enkakuji, and the Tennoji, which were family temples for the Shō family (Tōma Shi’ichi 1986, 305). Buddhist sutras were given to the Tenkaiji by the Korean court in the mid-fifteenth century AD (China Teikan 2008, 86).
Partial excavation of the Tenkaiji (Naha Shi 2000; Yamamoto Masa’aki 2003), built by King Shō Taikyu from AD 1450 to 1456, yielded the remains of part of a village found beneath the temple features, confirming that there were settlements in the area at the time of the construction of Shuri Gusuku (Miyagi and Tamashiro 2005, 177). In the mid-fifteenth to early sixteenth centuries AD, there was a main structure with two lateral extending corridors, a separate rear building for the priests quarters (hōjo), and a main gate and side buildings. The powerful Ōuchi daimyo family of northern Kyushu maintained close connections with the Tenkaiji, using its power to advance their attempts to establish diplomatic and commercial relations with the Ming court. In AD 1530 the Ōuchi gained the rights to manage Japanese tribute missions to Ming, which gave them an advantageous commercial position in relation to China and other countries in East Asia. However, they later lost their power, and the Shimazu became preeminent (Itō Kōji 2008, 89).
Construction of the Enkakuji was completed in AD 1496, and it was rebuilt in 1588. Excavations from 1997 to 2002 located all of the main buildings. It contained a rectangular pond crossed by a bridge, a main gate at the top of a flight of stone stairs, a main Buddha hall, a side building for the royal memorial tablets, priests’ living quarters, a garden, and a Lion Cave area (Okinawa Kenritsu Maizō 2002a).
The establishment of Buddhist temples in Shuri typified the complexity of relationships linking the Ryukyu Kingdom to the Japanese main islands. Who bore the costs of construction of these temples? What was the composition of artisan groups who designed and built these complexes? Sakihara (1987, 170) mentions that the Chinese envoy Chen Kan found in AD 1534 that the Enkakuji was closed to Okinawan local people. Perhaps it was open only on certain days of the year. It seems that the main function of these temples was to legitimize Chūzan sovereignty within East Asia. Although there is less archaeological evidence (p.262) compared to the remains of Zen institutions, Shingon or Esoteric Buddhism also played an early, important role in the development of Ryukyu cosmology and state ideology, being particularly important in the reign of Shō Shin (AD 1477 to 1526) (Smits 2000).
Shrines to Confucius and to the Heavenly Consort (Tian Hou in Chinese) were constructed in the Chinese port community of Kume but not near Shuri Castle. These were linked to trade but not directly to royal authority.
The main road leading from the castle to the port of Naha passed by the Sogenji temple, which housed the royal memorial tablets, on its way to the port community of Tomari and to the Chōkotei, a raised causeway about 1 km long that led to the island of Naha (Uezato 2009). The port community of Naha was located on a low island (Ukishima) with brackish estuaries to the north and south. To the north was the port of Tomari, the port for boats traveling to offshore islands such as Kumejima. On the southern side of the island was the main port for foreign shipping. Excavations have located a small triangular island (Watarichi Shima) with a cluster of buildings (Watarichi Mura) immediately inside the port on the north side and separated from Ukishima by a slip that was the berth for ships from China (Okinawa Kenritsu Maizō 2007c; Seto 2010b, 48). This small island was across from Omono Gusuku, a walled warehouse sitting on another small island in the middle of the channel to the inner harbor. On Watarichi Shima there was a warehouse for storing sulfur and separate warehouses for tributary goods from Miyako and Yaeyama. A trench 6 m × 62 m, dug to below sea level, yielded a large deposit of fourteenth to fifteenth century AD Chinese celadon, early remains of iron and copper working, and some 4,970 roof tile fragments, 2 percent of which were of the Koryo or Yamato Types of the Gusuku Period, indicating the presence of some kind of official structure at an early date. The rest of the tiles dated to the Early and Late Ryukyu Kingdom Periods. A total of 464 coins were recovered, including one specimen of the mysterious Ryukyu coin Taisei Tsūhō, said to be minted in AD 1454 (Okinawa Kenritsu Maizō 2007c, 29). The two distinctive characters for Taisei stand out sharply compared to the two characters for Tsūhō. The authors of the report conclude that coinage in general may have circulated in villages near the port (ibid., 188).
In the sixteenth century AD the entrance to the harbor was protected by two long arms, with the fortifications of the Omono Gusuku and Mie Gusuku at their tips. Nearby Watarichi Mura was a stone-faced platform from which officials and others could greet ships entering and leaving the harbor. From that (p.263) point the north arm included the Oki Gū Rinkaiji temple and two fortifications, Naka Mie Gusuku and Mie Gusuku. Tana (2010) notes that fresh water flowing from three convergent small rivers must have kept the inner harbor free from coral. An interior harbor protected from the open sea would have been easier to defend, resembling ports visited by Ryukyu traders such as Fuzhou, Quanzhou, and Ayuthaya.
At present, tidal flats separating the island of Naha from the mainland have long since been filled in, and there is a continuous urban zone extending from the edge of the East China Sea to Shuri and beyond. Prehistoric sites such as the Sachihijah Shellmound and Nazakibaru Site of the Gusuku Period show that the port area was an important region in prehistory. Along the Ajagawa River on the northern edge of present Naha, many sites yield kamuiyaki, whitewares of the Birosuku and Nakijin Types, and incised-flower-type celadon bowls, indicating that the antecedents of Naha port may lie in this inner area. Kin (2008a) notes that investigations around the present Naha port have not yielded materials from the twelfth to fourteenth centuries AD.
The “floating island” of Ukishima contained the Chinese residential area of Kumemura and the Japanese area of Wakasa Machi (Uezato 2008; 2009) (Figure 9.5). Kumemura was a residential walled compound for the Chinese, whereas the Japanese lived among the Okinawans. On the island were various religious and commercial buildings that served the trading community. The Oyamise and the Tenshikan were the center of trading activities with China. There were also religious institutions such as the Daoist Tensonbyo, Shintō shrines, mostly of the Kumano Gongen Sect (Uezato 2009, 74), and Buddhist temples. A number of Buddhist temples were located in the port and waterfront areas (Uezato 2009). A temple to Confucius was built in the seventeenth century (Itō Yoji 2010) near the cliffside temple of Naminoue, and places devoted to Mazu (Tian Hou), the Heavenly Protectress of sea travelers, were built in Naha, Naminoue, and Kumemura (Takahashi Yasuo 2010).
The Wakuta kilns at nearby Izumizaki appear to have operated in the latter half of the sixteenth century AD. They produced roof and paving tiles for Shuri Gusuku. The structure of the kilns is different from the general subterranean kilns of Kyushu and Honshu, being a single horizontal semisubterranean chamber. The ceiling portion has not been preserved, but from what remains it seems to be of a type not seen in south China but only seen in the area around Peking, indicating that Chinese craftsmen migrated directly to Okinawa from the Ming capital (Morimoto 2004). The tiles made there were stamped with the character dai sitting in a circle. Roof tiles and paving tiles for Shuri were made there. Ishii (2008, 90) noted that round eaves tiles from Locality R of Wakuta resemble Song dynasty examples found in Ningbo, Zhejiang. (p.264)
1. Naminoue Kumano Gongen, Gokokuji
2. Tenmangū, Chōrakuji
3. Wakasa Machi Ebisu Den
4. Wakasa Machi Jizōdō
5. Ukishima Jinja Chōjuji
6. Tenpi Gū
9. Naha Jizōdō
10. Ebisu Den
12. Oki Gū, Rinkaiji
13. Wakuta Jizōdō
14. Ameku Gū, Seigenji
15. Tenson Byō
16. Omono (Mimono) Gusuku
17. Mie Gusuku
18. Yarazamori Gusuku
The Expansion of the Ryukyu Kingdom into the Sakishima Islands
Archaeology provides a picture of independent agricultural communities in contact with traders from the north, bringing kamuiyaki and soapstone cauldrons, and from Fujian, bringing whitewares and celadons, in the period from the twelfth to fifteenth centuries AD. Elite burials indicate that chiefs were buried in distinctive stone cist chambers. A period of prosperous trade is attested by abundant trade ceramics. Evidence of the historically documented (p.265) conquest of Chūzan at the beginning of the sixteenth century AD is scarce, consisting of administrative sites and warehouses.
While Chūzan was consolidating its power in the Okinawa Islands in the fourteenth and early fifteenth centuries AD, it was beginning its expansion into Sakishima. The Miyako Islands first came under its control. Relations were first established with Miyako at the end of the fourteenth century AD. In the late fourteenth and early fifteenth centuries AD, Miyako chiefs offered allegiance to Chūzan and started sporadic raids on Yaeyama. They may have been somewhat unified by AD 1390, when the leader Nakasone Toyomioya pledged allegiance to King Satto. They must have relied on the Yaeyama Islands for timber, stone tool material, wild boar, and other products. Different factions within Yaeyama received support from Chūzan and Miyako, finally defeating the local lord Ōyake Akahachi in AD 1500. This brought an end to a period of independence and prosperity for the Yaeyama Islands from AD 1200 to 1500, when islanders were actively engaged in private trade with China (Ōhama 2005), indicated by abundant Chinese trade ceramics, particularly celadon. According to Ōhama, the Ryukyu Kingdom invaded in order to gain control of overseas trade and profit from it, and to control piracy. Local products traded to China include shells, marine products, medicinal plants, and woven hemp (Ōhama 2009). Ōhama proposes that Chinese cargoes were assembled in Fujian and offloaded by small boats in the Sakishima Islands. Profits from the private trade with Fujian may have attracted migrants from the Northern Ryukyus, since the total number of sites increased in the fourteenth and fifteenth centuries AD. The six ancestral shrines of Taketomi Island, thought to date from this period, are devoted to deities from Yakushima, Okinawajima, Kumejima, and Tokunoshima, indicating population movements from the north (Kin 2008b, 49).
For Miyako and Ishigaki, I describe below villages that flourished during the period of independence from the twelfth to fifteenth centuries AD and sites associated with local leaders around AD 1500, when the Ryukyu Kingdom consolidated its power over Sakishima.
In the archaeological sequence of Yaeyama, the period in which these developments occurred is known as the Nakamori Period, fourteenth to seventeenth centuries AD. Ethnohistorical sources describe a network of simple chiefdoms. Various historical sources shed light on Miyako and Yaeyama during this period, the most useful being the account of a group of shipwrecked Koreans who spent twenty-nine months on eight islands. The Koreans were repatriated to Korea via Hakata in AD 1479 (Iha 1974; Pearson 2003).
Archaeologically the most diagnostic artifacts are soft red earthenware wide-mouth jars with external lugs, found along with abundant Yuan, Ming, and rare Southeast Asian trade ceramics. Stone tools were replaced by iron tools, and the (p.266) presence of slag and tuyeres indicates ironworking. Cultivation of rice, millet, and barley is thought to have begun around AD 1200. Burials and village sites give a picture of life in this period.
Village sites dating from the twelfth to sixteenth centuries have ground-level houses (not pit houses) with paved floors as well as elevated storage houses. Many are located inland behind protective coastal vegetation. They yield the usual suite of cereal grains and bones of fish, cow, horse, goat, pig, and chicken.
Imposing stone tombs in both the Miyako and Yaeyama Islands have been linked, by writers such as Inamura Kenpu (1972), Kaneko (1964), and Tōma Shi’ichi and Nakama (1983), to the chiefs of the sixteenth century AD who came under the control of the Ryukyu Kingdom. The most elaborate examples, termed miyāka, are found in the Miyako Islands, in Hisamatsu (Matsubara), Irabu, and Hirara. The Hisamatsu examples are large stone cists sitting in or on a terrace enclosed by a large row of upright stone slabs (Kaneko 1964, 122–133). Inamura (1962) provides an account of an excavation in AD 1897 of an unnamed tomb. It was covered with soil and tree roots at the time of its discovery. It yielded a stone adze, hammerstone, stone weight, and other stone artifacts. Reddish brown earthenware vessels, of wide- and narrow-mouthed types, and perforated handles were found on top of the coffin lid. Inamura estimated that it would take six individuals to move the coffin lid into place (1962).
The most imposing of the chiefly tombs on Miyakojima is the tomb of Nakasone Toyomioya, who led the Chūzan forces of the Ryukyu Kingdom against Akahachi of Yaeyama and later became the leader (kashira) of Miyakojima (Sunagawa 1983). Built against a steep natural slope, it has two interior chambers, front and back, which are still used by the descendants for burial. The facade consists of seventeen cut limestone steps, and there are seven standing stones on the roof. Nakasone Toyomioya must have lived at the Sumiya Site. His father-in-law, the chief Akagi Tatoya, is said to have been buried at the Hisagai Tomb Site in Hisamatsu. On Taramajima, there is the tomb of Untabaru Toyomioya, a retainer of Nakasone Toyomioya, and his wife, who assisted in the subjugation of Chief Akahachi of Yaeyama. His family genealogy states that he became chief of Taramajima and Yonagunijima (Nakama 1983; Ōyama 1983). Two house- or casket-shaped stone cists with sloping roof are surrounded by a limestone wall 70 cm high with an arched gate. None of the recently investigated large elite burials in the Miyako or Yaeyama Islands dating from the fourteenth to sixteenth centuries AD contain grave goods; later burials (p.267) often had Japanese, Okinawan, or Fujianese ceramics placed on top or in front of them.
Sumiya Site, Miyakojima
The Sumiya Site has been the administrative center of the Miyako Islands since the fifteenth century AD, first as a chiefly settlement in the fifteenth and sixteenth centuries, when Nakasone Toyomioya lived there. It was the site of the local administrative center of the Ryukyu Kingdom from the early sixteenth to the late nineteenth century AD and became the administrative center for the amalgamated city of Miyakojima, established in 2006. A modern city hall stands to one side of the site, and part of the area that was excavated comprised the Ogden Building for the city Social Welfare Division under the US Civil Administration from AD 1945 to 1972. A large-scale excavation of about 3,300 sq m, consisting of excavation units of 4 m × 4 m, was undertaken for eleven months beginning in April 1990, following an earlier excavation in 1982 (Okinawa Ken Hirara Shi 1983; 1992; Hirara Shi 1999). Analysis of ceramics from the site shows a strong presence of trade ceramics in the fourteenth and fifteenth centuries AD (celadon bowls), fifteenth and sixteenth centuries (blue and white bowls), and seventeenth and eighteenth centuries (Fujian blue and white bowls and Satsuma ceramics) (Pearson 2003).
Judging from numerous finds of kamuiyaki, soapstone cauldrons, and Tong’an celadon, occupation of Sumiya and a number of other sites began in the twelfth century AD. Sumiya yielded many features such as building foundations and pavements, postholes, storage and refuse pits, and burials, the interpretation of which has been complicated by extensive disturbance since 1945. Features were divided into the following types: structures, stone arrangements, postholes, pits, hearths, and burials. The relationship of features to the stratigraphy is not clear in the report. Remains of at least twelve structures were recovered, but only two (Nos. 1 and 2) have been dated conclusively by associated artifacts to the fourteenth century AD.
Pit dwellings of an early type (fourteenth to first half of fifteenth century AD) were followed by buildings built on the ground surface (late fifteenth to early seventeenth century AD) and buildings supported by pillars or posts (Takemoto and Asato 1993, 227–229). Structures Nos. 3, 4, and 5 are possibly administrative buildings dating to the seventeenth century AD or later, and No. 9 is also a large building showing successive construction using posts set in holes. No. 4 with its postulated tile roof may be elite or administrative. My interpretation is that the large structures date to the period after AD 1500 when the Ryukyu Kingdom (p.268) assumed control. Other architectural remains consist of rectilinear areas paved with rough coral chunks. Children’s burials, dating to the sixteenth to seventeenth century AD, have been found under these pavements, the remains placed in rough rectangular box-like cists made of chunks of coral limestone. These cist burials are thought to date to the sixteenth and seventeenth centuries AD. Pavement dimensions are roughly 4 m × 4 m. They may be house floors, with child burials beneath the floor. Postholes are scattered through the site, one group in the northwest appearing to be the remains of an elevated granary. A total of 64 pits were dated by associated tradeware sherds; of these, 17 contained sherds dating to the sixteenth century AD or earlier, while 7 contained sherds dating to the sixteenth to nineteenth centuries AD. The others have sherds from both periods, the result of continuous occupation and extensive disturbance.
Analysis of shellfish remains from the site showed that in the lower layers, dating to the fifteenth century AD, shells from the outer ocean and from coral reef areas, such as Strombus luhuanus and Turbo argyrostoma, were abundant (Hirara Shi 1999, 297–298). Layer 2 and below are the layers of occupation that predate the arrival of officials from Shuri. Shells from this period include Tridacna, which could be used for shell weights, or Turbo shells, the operculi of which were used for tools. Shellfish meat was also eaten. Shells found in the upper layers, very large Turbo, Trochus niloticus, and Tridacna, were the remains of official feasts for Shuri representatives. The lower-layer shells are similar to those from other fortified sites, coming from inner bays, rocky areas, and tidal river mouths. There are no shells from mangrove areas or terrestrial shells.
Fragments of slag and iron cauldrons were found in the Ogden area, also in Layer 2. Some iron nails were found in the same areas as well as four knife blades, primarily from Layer 2, and two longer blades that could be swords. A few small fragments of iron slat armor were also noted. Metallurgical analysis of iron artifacts indicate that both iron sand and iron ore were used as raw material. Judging from relatively high percentages of manganese oxide, the raw material of some of the artifacts seems to have come from the region of northeast China and Korea. It is likely that the inhabitants used worn out iron cauldrons to make wrought iron and then worked the wrought iron into artifacts (Hirara Shi 1999, 267). There were also four bone spear points from Layers 2 and 3, thought to have been used for hunting or fishing. A review of trade ceramics at the Sumiya Site showed a small but significant number of very large and valuable trade ceramics, suggesting a center of political power (Pearson 2003).
At the beginning of the fifteenth century AD, the Yaeyama chief Ōyake Akahachi, whose residence was the site of Furusutobaru, was subjugated by the Ryukyu Kingdom. In AD 1522 and 1524 respectively, Ryukyu government warehouses for storing tribute were built in the Miyako and Yaeyama Islands. Local (p.269) textiles were a major tribute item sent from Sakishima to Shuri. It is interesting to note that a relatively high number of spindle whorls, fashioned from earthenware sherds, have been found in Sakishima sites of the fifteenth and sixteenth centuries AD (Uehara 2007, 260).
Sites in Yaeyama
The sequence in Sakishima, with particular reference to Yaeyama, is summarized in Table 4.3. Sites dating to the twelfth to thirteenth centuries in the Shinzato Subperiod of the Suku Period include Birosuku, which yielded early whitewares, kamuiyaki, and fragments of steatite cauldrons.
The centrally located, low, well-drained island of Taketomijima was an important center for Yaeyama, from which villagers commuted to agricultural plots on the wet, malarial islands of Iriomotejima and Ishigakijima, which also supplied timber and raw materials for stone tools. A zone of ancient settlement on Taketomijima, Shinzatomura, has east and west components. Shinzatomura East is contemporaneous with Birosuku and has yielded very interesting Chinese white and brown wares (Okinawa Ken 1990b). The fourteenth to fifteenth century AD Shinzatomura West and Hanasuku Sites have both undergone extensive excavation. The Shinzatomura West Site, Taketomijima, shows the distinctive clustered settlement pattern of old villages in the Yaeyama Islands. Such a pattern is not known in the Okinawa Islands. House sites are surrounded by stone walls with only a single narrow entrance to each compound, which was connected to other compounds. There was no central arterial road, although there was a central plazalike space surrounded by compounds. Several house sites contained more than one dwelling. Seventeen house sites were recorded, and four were excavated. The same type of clustered domestic plan of up to nine multiple contemporary dwellings was recorded on Taketomijima in the 1880s AD (Kin 2004). The site of Hanasuku lies on a ridge about 15 m above sea level, with defensive walls as high as 4 m (Nakamori 1999; Okinawa Ken 1994, 29–32; Ono Masatoshi 1997a; Ono 1999, 37–54). Its dimensions are about 500 m east-west, 200 m north-south, containing about forty clustered dwelling sites in bilaterally symmetrical arrangement. In each half, Hanasuku and Kumara, there is a large house in a single enclosure 30 m × 30 m surrounded by seven or eight medium and small dwelling areas comprising a block 50 m × 60 m. Around these are more enclosed dwelling sites, and each of the major blocks has an ancestral shrine (utaki) attached to its corner.
There are also a number of beaches in Yaeyama with dense scatters of trade ceramics. George Kerr termed some of these “trading beaches” (Kamei 1982, 117). He proposed that these beaches may have been visited by itinerant trading (p.270) vessels, that there may have been storehouses nearby, and that the sherds on the beach were from discarded broken wares. The beach at Komi and the offshore island of Pinishijima, Iriomotejima, are good examples. The ceramics found there have been dated by Kamei (1982, 131) to early Ming (fifteenth century AD). The beach at Shitadaru (Kudo), Ishigakijima, may have been the scene of a wreck, since no adjacent storehouse or settlement has been found (Ōhama 1994). Kamei (1982, 135) dated its ceramics to early and middle Ming, and Ōhama (2009) has dated it to AD 1450 to 1500. This is one of the largest beach finds of trade ceramics sherds in Japan, yielding a huge quantity of celadon, brownwares, and whitewares. Stylistic dating of ceramics from these sites indicates that the vigorous trade took place at the time of the heroic leaders, before the consolidation of power by Shuri.
Remains of a large, thriving village dating to the fifteenth and sixteenth centuries AD underlie the shoreline areas of Ishigaki City along the southern coast of Ishigakijima, designated the Ishigaki Shellmound, which was originally a sand dune backed by a low coral limestone terrace and facing a rich reef (Ishigaki Shi 2009). This settlement must have been occupied during the time of subjugation by the Ryukyu Kingdom. Analysis of human burials showed traits of mainland Japanese such as pronounced facial flatness, lending support to the proposal that there was immigration into the Ryukyus as far south as Sakishima at the beginning of the Gusuku Period (Ōhama 2005). In the same general area as the Ishigaki Shellmound, on a sand dune about 3.5 m above sea level, the sixteenth century AD warehouse site of the Chūzan Kingdom has been excavated. Although the site is disturbed, pillar bases for supporting a large building, in association with fifteenth century AD celadon sherds and later Okinawa ceramics, were recovered, along with stone and bone artifacts similar to those from Birosuku and Furusutobaru (Ishigaki Shi 1977). The zone of fifteenth to sixteenth century AD settlement on the old shoreline of Ishigaki City includes the adjacent sites of Hirakawa and Kidamori (Okinawa Kenritsu Maizō 2011). The Ishigaki sites do not show the same walled form as the sites on Taketomi Island. The walls may have been destroyed by the Meiwa Tsunami of AD 1771 or by subsequent construction.
As Sumiya was the residence of Chief Nakasone of Miyako, Furusutobaru was the center for Chief Ōyake Akahachi of Ishigakijima (Ishigaki Shi 1977; 1984; 1991; Tōma 1983; Takemoto and Asato 1993, 253–254; Okinawa Ken 1994, 8–13; Shimoji Kazuhiro 1999). The site is situated on a limestone ridge overlooking Miyara Bay (Figure 9.6). Fifteen walled enclosures have been identified. Enclosures (p.271)
1 through 4 are in a line with the seaward edge of the ridge, whereas the others form a row farther inland. Spaces of 190 m between Nos. 4 and 5, 40 m between Nos. 5 and 6, and 40 m between the two rows of enclosures may be the result of damage to the site. Enclosures 1 and 2 have been excavated. No. 2 had walls up to 2 m thick with a single entryway and central hearth. The large hearth was 130 cm × 85 cm and 10 cm deep. In No. 2 there were one hundred post molds, some containing stone wedges. No clear pattern could be interpreted. Enclosure 5 had an interior area of 600 sq m. The walls were 1.8 and 2.5 m thick. Enclosure No. 15 was a rectangle with length and width of 23 m and 20 m respectively. Burials with predominantly eighteenth century AD Tsuboya or Chinese ceramics are later than the main site occupation.
The Ryukyu Kingdom was a maritime entrepôt with economic and diplomatic links throughout East Asia. It arose between territorial powers in China, Korea, and Japan, with minimal defenses and limited agricultural productivity, but
(p.272) flourishing trade. The center of the kingdom was the Shuri capital, where the court formed a distinctive community with its own architecture and ritual program. Through its court and folk culture, language, and religion, it constituted a distinctive cultural community. It contained enclaves of traders and supported foreign and indigenous craftspersons with diverse specialties. Through military expansion to adjacent areas of the Ryukyus, the kingdom gained control over overseas trade and extracted tribute from them. The administrative capital of Shuri reflects the participation of the elite communities in various political and economic networks. The administrative capital and the commercial port were spatially separated and distinctive from the rural population. The kingdom expanded to include the island groups of Amami and Sakishima in addition to the offliers surrounding Okinawajima, and these were linked to Shuri through the extraction of tribute and a network of court officials. The kingdom was also linked to Japanese and Chinese religious and commercial communities.