Gaiso-Bo and Fuso Mortuary Practices and Human Migration in the Late Pleistocene in the Ryukyu Islands

Abstract

This chapter introduces and discusses the Pleistocene human migration from the view of mortuary practices in the Ryukyu Islands, such as gaiso-bo and fuso traditions, based on the late Pleistocene tomb site newly found in Ishigaki Island on the southern Ryukyu. Gaiso-bo is a traditional mortuary practice where rock shelters and caves are used as tombs, and remains of the deceased are exposed to the elements to undergo natural decomposition. Fuso mortuary practice where the body of the deceased, with soft tissues still remaining after death, is placed in an exposed environment on the surface over time to allow it to naturally decompose into bones. With the detailed introduction of these mortuary practices, I also discuss the Pleistocene mortuary practices in the Ryukyu Islands with comparative view from the cases of Pleistocene mortuary practices in other island environments like Wallacea and Near Oceania.

1. Introduction

In the Ryukyu Islands, there exists a burial tradition known as “gaiso-bo” where rock shelters and caves are used as tombs, and remains of the deceased are exposed to the elements to undergo natural decomposition [1]. Even after the bodies have already decomposed, they are not buried underground but instead left on the surface, which has been known as “fuso” mortuary practice in the Ryukyu Islands up to now. This chapter reviews and analyses those mortuary practices which reached their peak during the Ryukyu Kingdom period and became an established tradition. Some of these tombs have a distinctive shape known as “kameko-baka” or turtleback tombs, resembling ancient Chinese turtleback tombs. This reveals that the Ryukyu Islands has a unique culture that differs from mainland Japan, not only in the form of its tombs but also in the locations of these tombs and the traditional mortuary practices.

Prominent examples are the Mekaru Old Tombs (Figure 1A), located in Naha City from the Ryukyu Kingdom period [2]. It is a valuable archeological site where the Naha City Board of Education undertook excavations due to the development of Omoromachi area. Some parts of this tomb complex have been designated as a national historic site, allowing us to gain insight into the burial practices of the Ryukyu Kingdom period (13th—16th centuries).

Figure 1.

The appearance of “Fuso” tombs in the late medieval (Ryukyu kingdom period) and the placement of human remains in cinerary urns (Figure 1A and B taken from (2), C from (3)).

In the centre of the photograph stands a distinctive turtleback tomb, while on both sides, numerous graves carved into the cliffs are densely lined up. This view presents an unusual landscape not seen in other areas of current Japan, although it corresponds to the Edo period (17th–19th centuries) in Japanese history. The elaborate shapes of turtleback tombs are merely external features. It is inside the tombs where the unique Okinawan funeral tradition lives on (see Figure 1B). The entrance of the tomb on the cliff is located on the left side of the picture. As one enters the tomb from the cliff, you can see a single body placed inside. While the human form is well preserved, the skull has detached and rolled slightly away. The place where the body is exposed to the elements is called “shiruhirashi”. This body was not buried underground but was exposed to the elements on the ground.

In this “Fuso” mortuary practice, the body of the deceased, with soft tissues still remaining after death, is placed in an exposed environment on the surface over time to allow it to naturally decompose. Unlike most other traditional burials, it does not involve burying the body underground, and as a result, it cannot be referred to as “burial” in the conventional sense, thus we need to use the local term as “fuso” for such mortuary practice in the Ryukyu Islands. Hereby, the bones are ritually cleaned and placed in cinerary urns (referred to as “zokotsuki”, see Figure 1C) and are then stored on elevated shelves along the walls of the tomb. There are cinerary urns shaped like houses or jars lined up. Figure 1C depicts the interior of cinerary urns excavated from a fuso tomb called “Yacchinogama” in Kume Island [3]. There are two skulls within the cinerary urn. It is common for those cinerary urns to contain the bones of two to more individuals in the same cinerary urns, as a form of communal burial. It is evident that elaborate burial customs, including reorganizing existing remains when new deceased individuals are added, took place, signifying a form of repeated burials (multiple burials).

These tomb structures are not intended for use by individuals or a single cinerary urn; they are utilized by kinship groups and known as “monchu”. Inside tombs traditionally shared by the members of these kinship groups, the remains and cinerary urns of ancestors dating back to centuries are placed. When members of the modern kinship group pass away, they are placed in the same tomb alongside the remains of their ancestors. During the Ryukyu Kingdom period, such tombs were commonly used, and excavations uncovered thousands of fuso tombs, along with thousands of cinerary urns containing human bones.

The fuso tombs from the Ryukyu Kingdom period vary in structure, with some tombs merely consisting of caves or rock shelters with stone blocks covering the tomb entrance, while others feature structures resembling hafu (gables) and turtleback types. The deceased are exposed to the elements in an open space within the tomb, the shiruhirashi. After the bodies have decomposed, they are placed in cinerary urns or are communally put in a location called “ike”. Excavation research has revealed that in a single fuso tomb, there may be the remains of over a hundred individuals. These tombs exhibit several distinct characteristics:

1. They utilize natural rock shelters and caves.

2. Remains are placed on the ground, without being buried underground.

3. With the exception of specific cases, the human remains are typically not in an intact form but rather in a disarticulated state.

4. The exceptional cases are those in which human remains are found in shiruhirashi, usually in a well-preserved state.

This suggests that individuals who underwent “fuso” practice the most recently may not have needed reburial in a cinerary urn or ike for particular reasons. These unique tombs provide insights into the mortuary practices and customs of the Ryukyu Kingdom period; however, the recent excavation of Shiraho-Saonetabaru Cave Site on Ishigaki Island of the southern Ryukyu Islands (Figure 2) finds such “fuso” mortuary practice had been practiced during the late Pleistocene dated to around 27,000 BP [4, 5].

Figure 2.

Map of Ryukyu Islands and location of Shiraho-Saonetabaru cave site on Ishigaki Island (made and prepared by the author).

Based on this new finding, this chapter introduces these new findings and discusses the Pleistocene mortuary practices in the Ryukyu Islands with comparative view from the cases of Pleistocene mortuary practices in other island environments like Wallacea and Near Oceania. It also discusses the Pleistocene human migration from the view of mortuary practices in the Ryukyu Islands such as gaiso-bo and fuso traditions.

2. Pleistocene mortuary practice in the Ryukyu Islands

2.1 The Pleistocene case: shiraho-Saonetabaru cave site on Ishigaki Island

In previous studies, it is believed that the fuso and gaiso-bo culture date back to at least the late Jomon period in the Holocene. The question is, however, where its origin is. In 2009, during the survey of caves as part of the construction of the new Ishigaki Airport, human remains were discovered dating back to approximately 20,000 years ago from the Paleolithic [6]. This discovery garnered significant attention, as it was one of the oldest directly radiocarbon-dated human remains in the country. The site was named the Shiraho-Saonetabaru Cave Site and was explored by the Okinawa Prefectural Board of Education in collaboration with the Okinawa Prefectural Museum and Art Museum which had considerable experience in excavating cave sites. Large-scale and systematic excavation took place from 2010 to 2016 [4, 5, 7, 8].

The comprehensive investigations at the Shiraho-Saonetabaru Cave Site revealed human remains from various periods, including the late Ryukyu Kingdom period, Gusuku period, Shimotabaru period, incipient Jomon period, early Holocene, and the late Pleistocene. The discovery of human remains spanning over different cultural periods at a single site is rare and provides insights into the human usage of the cave over a long period, and has resulted in the designation of the site as a national historic site.

As mentioned earlier, while the presence of human activity at the site had been known, the mortuary practices and human remains from the incipient Jomon and Shimotabaru periods remained undiscovered. It was also revealed that there had been gaiso-bo practice during Shimotabaru period, which is significant information. Furthermore, there was an even more remarkable discovery, as the excavation from the late Pleistocene layers yielded substantial information that could help to reconstruct cultural activities based on the context of excavated human remains. This information can provide valuable insights into the character of the site and its evaluation, as deduced from the context of excavated human remains.

2.2 Pleistocene human remains and mortuary practice

During the late Pleistocene period, it is assumed that the cave had an opening on the southwest to south side [4]. The cave floor level was considered to be the highest at the opening area, descending in a stepped manner towards the northeast. Large limestone boulders, believed to have collapsed from the cave ceiling tens of thousands of years ago, are exposed throughout the area. Flat surfaces are formed by the soil and smaller limestone fragments fill the gaps between these boulders.

In this cave environment, human remains were discovered concentrated in five areas (“units”). The unit found at the highest point is Unit 1, and as the floor level descends, from Units 2 to 5 are numbered accordingly. Each unit contains disarticulated human bones, often encompassing the entire skeletal remains of multiple individuals, although individual identification is mostly not possible. This situation suggests that each unit was repeatedly and deliberately used as a burial site for human remains, implying that the entire cave may have been perceived as a burial ground.

Figure 3A depicts the condition of human remains of an individual when excavated (Shiraho Skeleton Number 4) from Zone H4 (Unit 4) of the Shiraho-Saonetadabaru Cave Site, which provides the oldest radiocarbon date within the site of approximately 27,000 years ago. Despite the initial appearance of disarticulation, the anatomical position of the bones is mostly preserved. An individual was laid in a manner where it was wedged into gaps among the rocks, with a posture estimated to be supine and flexed, both arms bent at the elbows with hands placed near the face (Figure 3B). The head orientation faces east to northeast. Typically, when a body is laid on the ground, the bones largely remain articulated and maintain anatomical relationships (human form) even after the flesh had decomposed. Skeleton Number 4 was preserved in a posture from their arrangement to be left there, albeit with slight disturbance in various parts. This suggests that the body decomposed and underwent ossification without being buried in the ground but left uncovered. The disruption in the bones is thought to have occurred due to the influence of gravity and terrain of the cave floor, as they ossified in the exposed space. In other words, it is inferred that the body was not buried in the ground but was placed in the rock shelter for fuso. This indicates the existence of a mortuary practice at the end of the Pleistocene which did not involve burying the body in the ground.

Figure 3.

Detected state of Shiraho skeleton number 4 and its reconstructed flexed posture for Fuso (taken from (4), edited by the author).

Before the discovery of Skeleton Number 4 in Grid H4, human remains (Skeleton Number 1) were found in Grid H6 (Unit 2), parallel along the cave wall with the heads and with both left and right femurs and pelvis bones in line. This arrangement suggested that, although comprising only a few parts, the skeletal remains belonged to one individual, maintaining anatomical positions. However, the distal and proximal ends of the left and right femurs were reversed, which puzzled the investigators at first. Yet, due to the strong flexed burial posture of Skeleton Number 4 unearthed later on, it became evident that Skeleton Number 1 was also interred in a tightly flexed burial posture and underwent fuso mortuary practice (Figure 4). Although it might be a coincidence, the head orientation is the same, facing east to northeast.

Figure 4.

Detected state of Shiraho No. 1 Skelton and gathered bones (taken from (7), edited by the author).

Furthermore, in this grid (H6), discovery of disarticulated human remains concentrated there from at least six individuals which were found at the feet of Skeleton Number 1. This situation, as mentioned previously, is similar to the mortuary practice of gaiso-bo, where previously laid human remains are gathered near the wall, and a new body is laid for fuso in the open space. From these observations, it can be assumed that fuso also took place in this grid, and that the human remains at the feet of Skeleton Number 1 and near the wall were likely the remains of individuals previously brought there for fuso and then gathered after having decomposed. Skeleton Number 1 was possibly the last body to be laid on the ground near the wall practicing fuso.

Multiple sets of disarticulated human remains were also found in three other units concentrated near the cave walls, with some of them partially maintaining anatomical positions. No sign of any habitation was observed inside the Shiraho-Saonetabaru Cave Site during the late Pleistocene, suggesting that the entire cave was utilized as a cemetery then, and that the origins of the cave use as tomb and fuso culture in the Ryukyu Islands can be possibly traced back to the late Pleistocene.

2.3 The composition of fuso burial at the site

The investigation of the numerous human remains recovered from this site provides insights into the cultural activities of late Pleistocene humans there. What is particularly intriguing is that among the minimum number of individuals totaling 19 found in this tomb, 18 were adults, both male and female, with only one being a juvenile. In Layer III B, which potentially contains human remains from the early Holocene, the minimum number of individuals (three) are also all adults [9].

In dental examinations, of the minimum number of 27 individuals spanning from Layer III B, containing human remains likely from the early Holocene, to Layer III C of the Paleolithic and below, only three juveniles were confirmed, including the one around 10 years old with deciduous molars and another around 15 years old with limited tooth wear [10]. From both skeletal and dental evidence, the presence of juveniles, especially infants, is exceptionally rare. This suggests that at this tomb, distinctions were made among adults, juveniles, and infants, with the possibility that infants were intentionally entombed in a different location.

3. Discussion-human migration and mortuary practices in the Ryukyu Islands

3.1 Tradition of Gaiso-Bo and Fuso practice in the Ryukyu Islands

Figure 5 shows the estimated time periods of representative gaiso-bo sites in the Ryukyu Islands. It is evident that these tombs, including the late Pleistocene case of the Shiraho-Saonetabaru Cave Site, span from the late Jomon period to the Yayoi and Heian parallel periods in Ryukyu. In most gaiso-bo sites, human remains are found disarticulated, with various body parts scattered, indicating a combination of gaiso-bo and fuso mortuary practices [11].

Figure 5.

Estimated time periods of representative gaiso-Bo sites in the Ryukyu Islands (made and prepared by the author).

The regional distribution of prehistoric gaiso-bo practice has been established through the burials at the Kinengenshi and Kinenkubansya Rock Shelter Tombs on Tokunoshima Island at the northern end to Shiraho-Saonetabaru Cave Site on Ishigaki Island in the south (Figure 6). They indicate that the practice of gaiso-bo during prehistoric times did not reach farther than the Tokara Structure Strait to the north, but extended beyond the Kerama Structure Trench which marks the boundary between the Okinawa Islands and the southern Miyako Islands, where cultural contact to each other was scarce in the late Pleistocene [11].

Figure 6.

Location of the major Gaiso-Bo sites in the Ryukyu Islands (map made and prepared by the author, and the picture (left above) taken from [12]).

Comparing the proportion of gaiso-bo to other burial practices in prehistoric Ryukyu Islands, it was revealed that gaiso-bo constituted approximately 27.5% of burials across the entire Ryukyu Islands. However, when comparing this to the area south of Tokunoshima, where gaiso-bo are at their northernmost limit, gaiso-bo account for as much as 38.6% [11]. This suggests that the practice of gaiso-bo was not a particularly unusual mortuary practice during prehistoric times in the Ryukyu Islands, especially considering the significant lack of rock shelter and cave excavation cases when compared to open-site excavations. Gaiso-bo, despite regional variations, are distributed across a wide area in a mosaic pattern, and they were not confined to a specific region or unique to the area [11].

Regarding the group composition at the Shiraho-Saonetabaru Cave Site, it became apparent that the majority of the late Pleistocene individuals were adults, with no infants or children among them. In terms of possible juveniles, only one individual out of a minimum of 19 individuals was identified, showing a clear bias [9].

Figure 7 illustrates a the age composition of skeleton in the major Gaiso-Bo sites in the Ryukyu Islands. In the previously discussed case of the Nakandakariyama No. 7 Tomb, which is gaiso-bo dated from the 16th to 17th centuries, 15 individuals out of a minimum of 47 individuals (32%) were juveniles [13]. At the Shiidachi Site of the Gushikawa-jima Site Group, gaiso-bo revealed that out of a minimum of 62 individuals, 24 of them (39%) were juveniles. Juveniles are present at a rate of 30–40%, and infants are also included in a certain proportion [14]. Even though the sizes of the groups in one tomb site differ significantly, the difference in age composition is evident.

Figure 7.

The age composition of skeleton in the major Gaiso-Bo sites in the Ryukyu Islands (made and prepared by the author).

Conversely, some gaiso-bo do not include juveniles. For example, the Shiidachi Site in Gushikawa-jima from the Jomon period, the human remains found in Layer 5B consisted of a minimum of 11 individuals, with only 2 juveniles (22%) included [15]. The 14 individuals from Layers 3 and 4, separated by an intermediate layer, were all adults, no juveniles, and certainly no infants [16]. In the gaiso-bo at the Ufudobaru Shell Mound, believed to date from the Yayoi to Heian periods, out of a minimum of 18 individuals, only one was juvenile (5%) [17]. Even during earlier prehistoric times, there are both, cases with and without distinction between young children and adults, including infants, and there is a noticeable imbalance. However, in the case of the late Pleistocene group at the Shiraho-Saonetabaru Cave Site, there is a high possibility of such distinction, which might provide insights into the origins of gaiso-bo and fuso mortuary culture.

The gaiso-bo and fuso mortuary culture of the Ryukyu Islands is believed to have had its beginnings in the Paleolithic during the late Pleistocene, continuing through the prehistoric Jomon period after the Holocene and potentially persisting as tradition until the Ryukyu Kingdom period, passing through the Gusuku period (medieval times). It should also be noted that various burial customs such as inhumation in the ground or cremation existed in the Ryukyu Islands during the prehistoric times as well as Gusuku periods. Crucially, the gaiso-bo and fuso mortuary culture persisted continuously from the prehistoric period. Eventually, it reached its peak during the Ryukyu Kingdom period, being referred to as “fuso” and becoming the mainstream of mortuary practices, despite the coexistence of other mortuary practices.

With the passage of time, the burials and mortuary practice in the Ryukyu became more complex, with segregating grave areas using stone structures, creating lateral graves in locations lacking caves or rock shelters, sealing grave entrances, and adding gables and turtleback-like upper structures (Figure 8). Accordingly, rituals such as bone-washing ceremonies for the bones having undergone fuso likely added to the complexity of these practices after the early modern times in the Ryukyu Islands. However, the question whether this kind of mortuary practice was specific to the Ryukyu Islands remains. Focusing on the island regions in Wallacea as well as Oceania during the Pleistocene may provide some answers, as those areas are similar to the Ryukyu Islands in the aspects of (1) island environment, (2) remoteness, and (3) tropic to subtropical climate.

Figure 8.

Chronological transition of rock shelter use for Gaiso-Bo and Fuso practice (made and prepared by the author).

3.2 Comparison of mortuary practice in maritime Asia and Oceania

In the Japanese archipelago as part of maritime East Asia, a number of Paleolithic sites appeared by around 40 ka on main islands (Honshu, Kyushu, and Shikoku Island). During the late Pleistocene, these islands were partly connected to each other by land-bridge before 12 ka and the distance of sea gap between the main island and Korean peninsula might have been less than 20 km [18, 19, 20]. On the other hand, the Ryukyu Islands (also known as the Nansei Islands), the southern islands in Japanese Archipelago, stretch with the length of about 1200 km located between Kyushu Island from Taiwan. They are basically divided as Northern Ryukyu, Central Ryukyu, and Southern Ryukyu groups, which also have numbers of late Pleistocene sites with human remains dated to around 35 to 20 ka (e.g. [18, 21, 22, 23, 24, 25, 26]).

Even during the late Pleistocene, there were sea gaps among these island groups. For example, there was a sea gap of around 200 km between Okinawa Island (current island area 1207 km²)15 in Central Ryukyu and Amami-Oshima island (current island area 712 km²) in Northern Ryukyu, while another gap known as Kerama Gap between Okinawa Island and Miyako Island (current island area 159 km²) in Southern Ryukyu was the longest with about 270 km.

The existence of several Pleistocene sites in Okinawa Island indicates that modern humans crossed such sea gaps to reach the island, though it is yet to be determined from which direction they migrated. Ishigaki Island is located around 130 km south of Miyako Island, while Taiwan Island is located about 130 km west of Ishigaki Island. The discovery of human bones dated into the Pleistocene and back to 28 ka on Shiraho-Saonetabru Cave Site is the first and the oldest evidence of human migration into Ishigaki Island by sea crossing over 100–150 km distance. Although the origin(s) of the early modern human in Ishigaki are yet unclear, the coast of Southern China including Taiwan which was part of the Chinese coastal region during the late Pleistocene, was the nearest location from Ishigaki. Among the Southern Ryukyu group, other Pleistocene sites are only located on Miyako Island, dated to between 26 and 20 ka [27]. The discovery of a Pleistocene site on Ishigaki also suggests a relationship between Ishigaki and Miyako Islands in the Southern Ryukyu.

In terms of mortuary practices by modern humans in Asia and Oceania during the late Pleistocene, there is still limited information available. Generally, most of the excavated human bones are fragmented and sometimes no detailed record of their archeological condition during excavation in tropical and subtropical environments like Southeast Asia and Oceania is provided. Thus, the exact processing method of burial or mortuary practice during Pleistocene is yet unclear (e.g. [28, 29, 30]).

For the cases in Mainland Asia, the complete modern human skeleton of Tianyuan man in China is one of the oldest burial examples in the late Pleistocene, and the body is thought to be buried [31]. In Southeast Asia, a modern human cranium from Tam Pa Ling Cave Site in Laos is dated to around 46 ka [32], while a modern human cranium from Niah Cave in Borneo Island [33] dated to around 40–45 ka, and another modern human cranium from Tabon Cave in Palawan Island dated to c. 16 ka [34, 35], both of which were part of Sunda land and the enlarged Asian landmass during the Pleistocene. However, all these remains are limited to the cranium and a secure determination of any potentially associated mortuary practice was not possible.

Regarding the cases in Sahul continent in Oceania, as well as the modern human skeletons from Willandra Lakes region in Australia, especially known as Mungo 1 and 3 (or WLH-1 and 3) are the oldest examples [36, 37]. Among them, Mungo 1 could be dated to between 30 and 20 ka by C14 dates and is the oldest case of secondary burial after cremation as some of the bones are burnt. Mungo 3 (or WLH-3) is an example of a flexed burial with ocher attached and dated to around 30 ka by C14 dates [37], though it was also dated to over 40–60 ka by optically stimulated luminescence (OSL) and U-Series dates [38, 39]. Although the exact date of the Mungo 3 burial has yet been controversial and unclear, the recent re-dating of shells associated with human remains and soil beneath the burial layer by OSL method shows the possible date of burial is around 32 ka [40].

In the Aru Islands which were part of the Sahul coast during the late Pleistocene, Leang Lemdubu Cave Site produces a fragmented female skeleton dated to around 17 ka [29], as well as Liang Nabulei Lisa excavation having also delivered human remains dated to the Terminal Pleistocene [28]. Those examples indicate that complex mortuary ceremony could be practiced for the burials. Except these cases, however, the number of archeological records of Pleistocene burial and mortuary traces is yet limited and thus remains unclear, even in the past continental regions both in Asia and Oceania. The most widespread burial traditions in Island Southeast Asia during the Early to Mid-Holocene appear to involve various forms of flexion, as well as other forms including flexed and decapitated, and seated burials. Those burials were rarely with grave goods and other material culture. Nevertheless, in the late Holocene, changes in burial practices as introduction of the grave goods into the burials started to emerge in Island Southeast Asia, even though there was continuity in the Terminal Pleistocene and Early Holocene tradition of flexed inhumation [41].

The number of such archeological evidence is much limited in the Pleistocene maritime Asia. In Wallacea, currently the oldest modern human remains are from the Tron Bon Lei rock-shelter in Alor Island which could be dated between 17 and 12 ka [42, 43, 44], see also the chapter by Sofia Samper Carro). Among the three human remains, a female skeleton associated with some shell-made fishhooks in Alor Island dated to around 12 ka by C14 dating [45]. High frequency of archeological finding of human burials in cave and rock-shelter sites could be caused by bias of taphonomy, yet all these evidences indicate cave and rock-shelter are one of the major cemetery locations for modern humans, as well as for archaic humans.

4. Conclusion and further remarks

In terms of complex mortuary practices by modern human in maritime Asia and Oceania, the fuso and gaiso-bo traditions of the Ryukyu Islands could be one of the major mortuary styles in these regions. Although it was the case in the past coastal region of Sahul continent, the Pleistocene Lemdubu burial in the Aru Island dated to 17 ka shows the modern human did have some kinds of complex mortuary traditions in the late Pleistocene. This continues in Southeast Asia and Wallacea into the early and mid-Holocene before burial practices in the region drastically change with extended interment and the addition of a variety of grave goods, likely connected to the upcoming of farmer societies [41]. Much later after the Neolithic times, the number of secondary burials with jar and pottery containers increased in Wallacea and Oceania. For example, Teouma cemetery site on Vanuatu, which is one of the earliest Lapita sites in Oceania possibly used by early Austronesian migrants with Neolithic culture, has produced over 68 mortuary contexts including several Lapita pottery enclosing human remains dated to c. 3000 cal. BP [30]. One of such cases is significant as the skull part was removed and re-buried with pottery as secondary burial. Although its number is yet limited, similar burial practices are also reported from the Neolithic burial sites in Taiwan [46, 47], which could also be used by early Austronesian groups.

The ethnographic studies also report many cases of such secondary burials and mortuary practices similar to fuso and gaiso-bo traditions in Island Southeast Asia. For example, the early twentieth century case from Central Sulawesi was reported, in which the corpse was placed nearby the house to decompose in the primary mortuary ceremony, after which the bones were transferred to a cave cemetery and placed on the surface with jars and accessories [48]. Among the Traja people in Southern Sulawesi, high-ranking dead individuals are placed inside the house for a certain period of time during which several magnificent ceremonies are held, and then finally transferred to a cave cemetery or niche tombs that are artificially dug into a natural lime stone cliff and are interred there with a variety of burial goods [49].

More various mortuary practices are also reported from Borneo (or Kalimantan) Island including the case of Murut people as corpses belonging to the chief class are placed for exposure on an artificial or natural table for about a year for the primary mortuary ceremony, before the skeleton is to be buried in a secondary mortuary ceremony. On the other hand, the dead among the Shinongo people are interred in a primary burial ceremony first, but will be unearthed later for cremation, and then the burnt bones are finally placed on top of the pile in a house [50]. A similar practice was reported from the early Holocene of Ille Cave in El Nido, northern Palawan, as the body had been systematically disarticulated and defleshed first before interment, followed by fragmentation of selected elements of the body when the bones were still fresh. The body was then burnt to varying degrees [51]. All these ethnographic and archeological cases show a great variety of complex mortuary practices, though they are basically very similar to the fuso and gaiso-bo mortuary traditions in the Ryukyu Islands.

Further comparative researches on how such complex mortuary traditions reported by ethnographic studies could be traced back in archeological and anthropological studies are our future tasks. However, our new finding at Shiraho-Saonetabaru Cave Site on Ishigaki Island indicates such complex mortuary practices known as fuso and gaiso-bo traditions could date back to the late Pleistocene and to at least 28 ka. Also, the significance of a very detailed recording of each anatomical location and condition of excavated human remains and restoration study of fragmented human remains should be emphasized for future studies in maritime Asia and Oceania, especially for reconstructing prehistoric fuso and gaiso-bo-like mortuary practices basically with fragmented and very fragile human remains.

Acknowledgments

This research is supported by a Japan Society for the Promotion of Science Research Grant (Grant Number: 16H06409, 20 K20504) as well as the research grant of NIHU Global Area Studies Program, Maritime Asia and Pacific Study Project.