Ancient bones in honor of the goddess Sinlaku: New findings on the island of Kosrae, Eastern Micronesia

Human remains were found on the island of Kosrae in a region formerly known as a place of worship for the ancient goddess Sinlaku. These burials were found to be quite unusual in comparison to previously documented burial patterns on the island. They provided the opportunity to study the changing ritual practices on the island around 300 BP (AD 1650). Historic ethnographic accounts by the German explorer, Sarfert, from 1908–1910 provide context for religious beliefs that are otherwise extremely difficult to assess within an archaeological context. The oral history of the island tells of the arrival of a foreign group to the island that may align with the onset of these new burial patterns.

Se hallaron restos humanos en la isla de Kosrae, en una región antiguamente conocida como un lugar de adoración de la diosa Sinlaku. En comparación con patrones funerarios documentados con anterioridad, estos enterratorios eran inusuales. Otorgaron la oportunidad de estudiar las cambiantes prácticas rituales de la isla alrededor del 300 AC (1650 DC). Los informes histórico-etnográficos del explorador Alemán Sarfert, en los años 1908–1910, sitúan el contexto de las creencias religiosas, que de otra manera serían extremadamente dificil de evaluar dentro de un contexto arqueológico. La historia oral de la isla habla sobre el arribo de un grupo extranjero que podría ajustarse con la aparición de estos nuevos maneras de enterramiento.

Introduction

Recent findings of human remains on the island of Kosrae in Eastern Micronesia provide evidence of the development of new ritual activities at AD 1650. These remains cam e from two distinct sites located 550 meters apart on the northern coast of the island (see Figure 1). The first sample came from a large deposit of human bone 150cm below the surface in sandy soils near the beach in Yekula. The second came from a burial within Wiya Bird Cave from which only a single skull was discovered. Despite the very different contexts of these deposits, they were found to be contemporaneous. The deposit at Yekula dated to 290 BP ± 25 years and the burial in Wiya Bird Cave dated to 286 BP ± 20 years.

On the island of Kosrae, there are references within the oral tradition to a post-settlement migration. Stories tell of a different group of people arriving that fought against those who were living there before. These newly-documented burial patterns show a marked difference to burial practices that preceded them and may represent the arrival of a new group of people.

Prehistory of Kosrae

Kosrae is the easternmost of three volcanic islands in Eastern Micronesia that are surrounded by a multitude of low-lying coral atolls. The earliest recorded dates for the settlement of Kosrae are roughly 1900 BP (Athens 1995). The earliest dates for the neighbouring island of Pohnpei, located 300 miles to the west, are also at 1900 BP from deposits found at Nan Madol (Athens 1990) and Ant Atoll (Galipaud 2001), except for a single date of 2100 BP from a shell on Ant Atoll, which may suffer from a marine reservoir effect. Additional dates of 1700 BP have recently been obtained from Mwoakilloa Atoll lying between Pohnpei and Kosrae (Poteate et al. 2015). The Marshall Islands to the east provide early dates at 2000 BP on Kwajalein (Beardsley 1994), 1900 BP on Majuro (Riley 1987), and 1800 BP on Utrok (Weisler 2001). Kiribati has a 2,000 year history as well (Thomas 2015). These dates thus exemplify a major wave of migration across the atolls and remote high islands of Eastern Micronesia roughly two thousand years ago.

Late Lapita plain-ware pottery is common at sites dating to this early time period and was found at sites on Pohnpei (Athens 1990; Galipaud 2001), Mwoakilloa (Poteate et al. 2015), and Kosrae (Athens 1995). The only recorded decoration has been thumb-impressed notching found on Ant Atoll (Galipaud 2001). The region as a whole therefore appears to have first been settled by a common group of people spreading across the region. Vanuatu and the Southeastern Solomon Islands have been suggested as the place of origin based on linguistic analysis (Rehg 1995). It is only at this time that a drawdown in sea level allowed the coral atolls in this region to re-emerge, providing a pathway of stepping stone islands for settlement (Dickinson 2001, 2003, 2004).

Figure 1.

Map showing the location of Yekula and Wiya Bird Cave on the island of Kosrae (background satellite image provided by Google).

The absence of food plants prior to their introduction by people would have limited these early settlers to small, nomadic populations. Studies of pollen cores conducted by Athens et al. (1996) have shown that a period of forest transformation took place from 2000 BP to 1500 BP, representing the establishment of permanent agriculture on Kosrae. The earliest megalithic sites first appear around 900 BP at Nefalil (Cordy et al. 1985), Likihnluhlwem (Bath et al. 1983), and Safonfok (Beardsley 2005) with the construction of the large megalithic site at Lelu beginning at 700 BP (Richards & Hobbs 2015). Athens (1995, 2007) discovered that a significantly greater number of radiocarbon dates have been found from this time period and after, representing a large increase in the population. Peterson (2006) has argued that this rise in population was fostered by the hybridization of two different breadfruit varieties that provided a reliable food source. The rising population and the construction of large megalithic sites are indicative of the formation of a complex chiefly system. The development of megalithic sites occurred at the same time or slightly before on the island of Pohnpei, where the large ceremonial center at Nan Madol has recently been dated to AD 1200 (Athens 2007).

To the south of Pohnpei and Kosrae are two atolls that are Polynesian outliers, Kapingamarangi and Nukuoro. Archaeological excavation has been performed on both islands within deep deposits on the main village islets. The earliest secure dates for both islands are roughly 350 BP (Davidson 1967, 1968, 1971; Leach & Ward 1981) suggesting that a later migration of Polynesian people occurred at this time.

The Worship of Sinlaku

The German explorer Sarfert (1919) documented the existence of a spiritual cult on the island of Kosrae that worshipped a goddess named Sinlaku (also known as Sinlaka). Sinlaku was the vegetation goddess and was worshipped at the time of harvesting.

“All fruits, especially the highly esteemed breadfruit stood under the protection of the vegetation goddess Sinlaka, who provided them with growth and prosperity. In her honour and in order to thank her a special breadfruit harvest feast was celebrated”

(Sarfert 1919:101).

The existence of the goddess Sinlaku is still known of today on the island of Kosrae, though it is most commonly referenced in opposition to Christianity, the common story being that the goddess Sinlaku ran away once the Christian god came to the island with the missionaries in the 1870s.

Sarfert documents the existence of two spiritual centers on the main island of Kosrae (formerly called Ualang) where the goddess Sinlaku was worshipped. One, he wrote, lay beneath the mountain Tanfokol on the southern coast of Ualang [Kosrae] which is believed to be the mountain currently called Finkol, the highest mountain on the island. This place has been identified as the site at Menke, which has undergone two field seasons of research by Beardsley (2012, 2014) and is believed to pre-date the construction of Lelu. Twelve house structures were located within the site, one of which is a very large mound that is believed to be the central temple where nine priests would worship, according to Sarfert. During the ceremonies that took place at the site, the priests lived there in seclusion.

According to Sarfert, Sinlaku was also worshiped at “Uia [Wiya] and Sialat.” More recent documentation of oral history included the story of a sorcerer named Sinlaku found at Yekula (Phillip 1963). Yekula lies between Wiya and Sialat and so it may be presumed that the area covered by all three place names may represent the place of worship. Religious practices were also said to take place at Loal where previous archaeological work has also been performed. Despite their common spiritual nature, each of these sites shows unique characteristics. Cordy (1983) recounted the description of a spirit place said to be located at Loal that was “a living site for priests dedicated to a certain spirit and a spot for ceremonies involving the spirit” (122), though the site was not relocated and therefore not described in detail. The site at Loal consists primarily of rectangular stone wall alignments surrounding living areas that are located at the edge of a mangrove swamp. Dates obtained at Loal [Lacl] of 530 BP (Welch et al. 1990) agree with its later development. Previous dates have been obtained from Wiya Bird Cave of 610 BP (Sinoto 1982) and the Wiya area of 548 BP (Cordy 1981) suggesting that this area was only settled later.

Thus, it appears that while Menke may have been occupied during the early prehistoric period, religious sites at Wiya and Loal developed later. This assessment is supported by Sarfert who writes, “Sinlaka built herself a house on the place Fal, at the foot of Tafonkol in the region of Fenkol [Menke]. Later on she left Fenkol and lived in Uia [Wiya]” (1920:396). Sarfert continues: “it also seems as if the cult was introduced in Uia [Wiya] later than the one in Fenkol [Menke]” (1920:399). Unlike Menke, there are no large stone structures or megalithic temples at Yekula marking the area where Sinlaku was worshipped. The only such features found on the surface are a few small areas of coral paving. Instead, the main feature of the site where Sinlaku was reportedly worshipped is a large scattering of oven (uhmw) stones amongst a variety of broken shells known to have been consumed as food. Additionally, surface investigations have found shell tools of different types as well as small stone adzes. Also found were many broken pieces of cowrie shell.

The story describes sorcerers at Yekula who could make food appear by magic (Phillip 1963). An earth oven (uhmw) would be prepared and a cup of sea water would be poured onto the earth oven, which would become sea food, and a cup of freshwater would be poured, which would become land food. These stories are similar to descriptions for rituals described by Luelen Bernardt (1977) that took place in Pohnpei at the Pahn Takai rock shelter. During these rituals, an earth oven would be prepared and a person would be placed inside. When the earth oven was opened, the person would be unharmed and food would be there in their place that had been provided by the spirits.

Findings

The Yekula burial

In February of 2014, human bones were brought to the Kosrae police station. They had been found at a construction site within sand that had been mined on island. It became immediately apparent that the bones most likely originated from a disturbed burial and the location of the sand mining at Yekula was investigated. Bones still lay at the bottom of the pit and in the profile a skull was partially exposed. A salvage excavation was then performed for fear that the deposit would be further disturbed. The excavation revealed two broken skulls nearest the mining activity and an intact skull. Beneath these skulls was a jumble of bones, including another cranium packed within 15cm of soil. There was no single well-defined burial and the deposit appeared to represent an open pit in which the bones of several individuals were placed within a secondary context. The whole deposit of four individuals took up less than one square meter. Contained within this burial was a shell arm band and fine shell beads (see Figure 2).

It was recommended that no further excavation take place at the site until a controlled archaeological investigation could be performed. Despite these recommendations, sand mining continued to take place and monitoring was performed. Eight pieces of human bone were discovered during monitoring. This material appears to have been remains from prehistoric disturbance of the burial when it was an open pit. They were clearly associated with the burial, but as they were in a secondary context, they were chosen for destructive analysis including ancient DNA analysis and radiocarbon dating.

In November of 2014, sand mining resumed at the site despite recommendations against it. Further mining activities uncovered a buried earth oven with a diameter of at least one square meter located 10 meters from the bones at the same depth. It is unusual for a cooking site to be located so close to a burial site. A sample was taken from the earth oven, but as yet radiocarbon dating has not been performed.

Figure 2.

Shell arm band and shell beads found with human remains at Yekula.

At this same time, additional deposits were found containing a large amount of human bone. However, the excavator refused to stop to allow a study of the burials and continued mining a few meters from the deposit. Quick observations were made as the bucket of the excavator continued overhead which showed the burial to appear almost exactly the same as the one before. A sample of the human remains was collected for fear that they would be destroyed and the ability to make collections in situ would be lost. Recommendations were made to the landowner, the excavator, and the Kosrae Historic Preservation Office to cover the burials and leave them intact for future archaeological investigation. It appears that this advice may have been heeded, but the preservation status of the burials can only be known once an investigation is performed.

An initial morphological assessment of the bones showed that there was a high degree of breakage, particularly of long bones. This breakage is not associated with the disturbance of the heavy machinery, as the bones had been carefully removed, but is believed to have occurred prior to burial. A mandible also showed an abscess that appeared to have healed over. Further analysis is currently being performed.

The Wiya Bird Cave burial

Wiya Bird Cave lies 550 meters to the west of the Yekula site. In July 2014, it came to the attention of archaeologists that a skull had been found in the cave during digging activities by the landowner to obtain nutrient-rich guano soils to sell to farmers. Japanese migrants are believed to have been the first to perform such activities in the 1920s. It seemed likely that any burials contained within the cave had been disturbed long ago. A large natural pond exists within the cave that forms from water dripping from the cave’s ceiling (see Figure 3). A proper excavation of the cave would be very difficult, requiring the drainage of the cave using water pumps. The skull and mandible were donated to the Kosrae Historic Preservation Office by the landowner, Dyson Albert. The skull was fused together and was in good condition.

Previous archaeological excavation was performed by Sinoto (1982) at the mouth of Wiya Bird Cave within layers surrounding stone alignments. The only pre-European artifact found was a grinding stone, however, buried midden deposits and a deeply buried earth oven provided evidence of a prehistoric occupation. Charcoal from the earth oven dated to AD 1340 ± 85 (Beta 1173) while charcoal from layers near the stone platform dated to AD 1735 (Beta 1172). These dates thus provide endpoints for the occupation of the site from 215–610 BP. Within the oral history of Kosrae, the cave is associated with a large eel whose travels formed many of the mangrove channels (Sarfert 1920).

Figure 3.

Photograph from inside Wiya Bird Cave looking out.

Cave burials have been seen elsewhere in the Pacific Islands. A cave burial dated to this same timeframe was also discovered in Hawai‘i on the island of O‘ahu at the Halawa rock shelter (Langenwalter & Meeker 2015). Here, skull fragments and teeth were discovered in 1966. Though the human remains were not dated, the stratigraphic context spanned a time frame from the 14th to 19th centuries AD, similar to Wiya Bird Cave. Cave burials found in Palau in Western Micronesia date to a much earlier time (Fitzpatrick 2002), roughly 2000 years BP.

Analysis

Radiocarbon dating of material

A single sample from each site was radiocarbon dated. A piece of cranium from the initial burial found at Yekula was sent to the Keck Carbon Cycle AMS facility of the Earth System Department at the University of California in Irvine. It was dated to 290 +/−25 BP (PSU-6368). A tooth from the skull found at Wiya Bird Cave was sent to the radiocarbon dating laboratory at the University of Waikato. It was dated to 286 +/−20 BP (Wk-40161) with a calibrated date range of AD 1650–1680 (see Table 1).

Dental analysis of Yekula material

Forty-four (44) teeth from the initial Yekula burial were sent Dr. Rebecca Kinaston at the University of Otago for morphological analysis. Dr. Kinaston found that the teeth represented at least four individuals: a 6–7 year old child, an 11–13 year old child, and at least two adults. There was very mild wear on the teeth with no evidence of cavities and very little calculus (calcified plaque). Some linear enamel hypoplasia (LEH) was noted, which is indicative of some type of stress during childhood. It is likely that the people were eating soft foods, such as starchy staples. The 11–13 year old child had very crowded teeth. The adult displayed a Carabelli’s cusp which is a non-metric dental trait associated with certain populations.

Table 1.

Radiocarbon dates.

Kosraean burial practices

The most well-know and most visited burial site on the island of Kosrae is at the megalithic site of Lelu, the remains of an ancient city where the royal lineages lived, including the paramount chief, tokosra, until the mid-19th century. Within the ruins of the city are five stepped-platform burial chambers composed largely of coral and stone. Recent uranium-thorium dating of coral from the tombs has placed their earliest construction at AD 1310 (Richards & Hobbs 2011).

Gulick (1862) describes the burial practice for the royalty at Lelu as a double burial. The tombs provided only a temporary burial location until the flesh had decayed. Afterwards, the remains were disinterred and placed in the blue hole within Lelu Lagoon, where a small man-made islet called Yenasr served as the altar for the final funeral rites (Cordy 1993). In 1910, the first documented excavation of a tomb took place (Sarfert 1919; Cordy 1993). Of the remains that were found, only four bones and some teeth were identified to be human, believed to be a man in his fifties. In addition, the bones of numerous birds and dogs, as well as Spondylus shell and pearl shell valuables were found and determined to have been included as grave goods. Cordy (1993) argued that the sparse nature of the remains and grave goods, including incomplete necklaces, fits with Gulick’s description of the burial practice. If the bodies were to be only buried in the tomb, one would expect to find a more complete human skeleton and more abundant grave goods.

Many similarities may be seen between Lelu and the site of Nan Madol on the island of Pohnpei three hundred miles to the west, most notably in the use of prismatic basalt in its construction and its function as a royal city (Bath 1984). However, there are also many distinctions, particularly the style of tomb. At Nan Madol, the tombs are shaped as rectangular platforms of prismatic basalt instead of truncated coral pyramids. Also, at Nan Madol, a far greater amount of human remains was found at the primary tomb complex on the islet of Nan Dowas during the 1963 Smithsonian Expedition led by Reisenberg, Evans, and Meggers (Rainbird 1994). While the results of this expedition have never been reported beyond three radiocarbon dates from the islet of Idehd (Long 1965), the author has personally viewed the huge amount of human bone, pearl shell fishhooks, Spondylus shell jewellery, and beads collected at that time, which is still held within the Smithsonian archives today. Hence, at Nan Madol, the double burial described by Gulick was not practiced at Nan Dowas. However, oral history accounts of the warrior Isokelekel, who conquered Nan Madol, provide two different locations for his burial: one at Peinkitel and one at sea (Hadley 1981). At Peinkitel, a tomb more closely resembling the stepped pyramid style at Lelu may be found and the two stories combined fit the double burial pattern. Interestingly, oral history accounts say that Isokelekel came from katau peidak, far upwind to the east.

Recently, dating of Nan Madol has placed its earliest construction at AD 1200, though the occupation of the two sites likely overlaps. Even within the short one hundred years separating the sites, burial practices appear to have changed quite a bit. Greater similarity may be seen between the burial structures at Lelu and stepped-pyramid tombs found in Polynesia, notably in Tonga on the island of Tongatapu. Clark and Reepmeyer (2014) document the presence of a three-tiered tomb for the paramount chief of the island, Tui Tonga, at Heketa that was constructed in the 14th century AD, around the same time as the construction of Lelu. These similarities may suggest connections between these island groups at that time.

Other sites on Kosrae contain burials within earth-filled platforms with stone facings such as at Tepat (Cordy 1981) and Lelu (Sarfert 1919). Cordy (1993) argued that none of these tombs could clearly be demonstrated to be older than AD 1860 and so no archaeological data exists for the non-royal tombs before that time. A historic tomb of more recent age was found during a survey of Yela Valley atop an artificial islet at the mouth of Yela River. This tomb was decorated with post-World War II glass bottles and was the grave of the landowner’s grandfather (Thompson 2013) showing that such burial practices extended into more recent times.

Of the European explorers who first visited Kosrae in the early 19th century, only Lesson (1839; Ritter & Ritter 1982) commented on native burial practices. Lesson was aboard the first scientific voyage to Kosrae with only a few recorded sightings of the island before then (Hezel 1979; Ehrlich 1981). Though later explorers would visit the royal residence on Lelu, only Lesson observed that the large mounds were in fact tombs. Of the common people of the main island of Oalang, he describes the burials as tombs beneath small open-sided huts within sugar cane and breadfruit trees high into the mountains. They were at times collected together in “little villages inhabited by the dead” where many relatives had been buried together. Under the roofs were found tools and looms that he believed to be the property of the dead. The fact that he uses the term ‘tombs’ to describe the commoner’s burial place would suggest that a tomb-like structure was being used prior to European influence, though how far back is uncertain.

In his survey of Wiya Bird Cave, Cordy (1981:250) first documented the practice of burying the dead in “pits in the sand in clustered localities & in the floor of the Wiya bird cave.” However, there is little more mention of the burials than this, and it appears that nowhere else on the island have the remains of such graves been found. In comparison to the royal tombs, the human remains unearthed at Yekula show no evidence of care given, or of respect for a proper burial. In fact, the opposite may be seen. The remains of several individuals are included together, which is indicative of an open pit, and most of the bones are broken.

Jones et al. (2012) provide an interdisciplinary approach for identifying cannibalism in Fiji, where ritualized cannibalism associated with endemic warfare was an inherent part of the ancient culture. Fijian warfare was driven by the need to acquire human sacrifices that would be ritually eaten (Clunie 1977). Borrowing from Turner (1983) and Turner & Turner (1992, 1999), they include deliberate bone breakage, cut marks, evidence of cooking, abrasions caused by anvils, and the crushing of vertebrae as osteological evidence of cannibalism. However, they admit that “cultural differences in mortuary practices coupled with undocumented and potentially highly variable methods of ritualistic cannibalism, precludes a strictly defined set of criteria for identifying this behaviour in archaeological contexts” (Jones et al. 2012:129). Regardless, it is clear that the human remains found at Yekula represent a distinctive burial context in comparison to the care bestowed on the royalty at Lelu. Jones et al. point out that there are many forms of cannibalism which may be practiced by one’s own people as an act of honor, as well as one’s enemies as a means of defilement. The broken nature of the bones found at Yekula may suggest some form of ritualized cannibalism occurred, especially given the oral history of the area as a sacred place.

The discovery of the complete skull at Wiya Bird Cave suggests something different still. The cave is a unique natural feature on the island of Kosrae. It is more closely likened to the deposition of human remains into the blue hole and may well represent a similar ritual activity.

An oral history of migration

In Kosrae, to this day, the Story of Niwa is still told. This story tells how a giant, or giants, came to Kosrae, killed and ate many people, forcing the island’s population to flee. All the royalty gathered into their canoes and sailed away. A poor woman named Niwa sat on a rock at the mouth of the harbor and begged these people to take her with them, but all the canoes were full and she was abandoned on her rock (Phillip 1963). Still today there is a rock at the mouth of Lelu Harbor called Niwa’s rock. Niwa hid in the mountains away from the giant(s) and raised her son.

Myths are difficult to accept as literal accounts of history. Yet, they represent the traditional history of a people passed down over centuries. Reduced to its most basic components, the story of Niwa tells of the arrival of a foreign group of people, more powerful than those that were living there at the time, who pillaged the villages, cannibalized their victims, yet allowed those people who were not in power to remain. This story is not so different from bouts of war that have occurred time and time again across the world.

On the island of Atafu, archaeological studies have shown a dramatic shift in diet from dog to pig occurring at roughly 300 BP (Addison et al. 2009). This shift in the population matches the oral history of Tokelau, which says that the current people living on island were not the original settlers. The first people to settle Tokelau were chased away by a foreign group of people who came with many canoes (Matagi Tokelau 1991). The story is similar to the story of Niwa told on Kosrae, with the exception being that on Kosrae, Niwa, and her descendants, remained on the island. Tokelau is a tiny atoll with a small population and no place to hide, while Kosrae is a large high island with a larger population and mountains in which to escape.

The radiocarbon dating of the human remains at Yekula and Wiya provide evidence of a cultural change occurring on Kosrae at the same time as the population shift occurred on Atafu. The Yekula burial appears to be a mass grave that may show evidence of cannibalism. At the same time, we see the settlement of Polynesian atolls to the south of Kosrae, Kapingamarangi and Nukuoro. While further study is needed, it appears that the Polynesian migration that settled these atolls may well have also reached Kosrae, leading to the establishment of new spiritual centers on the north coast of the island and the burials found there.

Discussion

While the historic ethnographic accounts of Sarfert and the oral history related to Yekula and Wiya make it apparent that religious rituals took place on the north coast of Kosrae, it is unclear how the burials relate to specific religious practices given the nature of their discoveries. The density of the bone deposits at Yekula appears to be indicative of a secondary burial of de-fleshed bones or multiple burials within an open pit. There are clearly many individuals buried within a small space. Wiya Bird Cave is one of the few large caves on the island and is thus of special significance.

What is certain is that the two burial deposits date to the same time period at roughly AD 1650, are representative of several individuals, and represent an unusual form of internment not seen in the island’s prehistory to date. The settlement of Polynesian outliers to the south of Kosrae at 350 BP represents a movement of Polynesian people into Eastern Micronesia. The replacement of a settled population by migrants in Tokelau at 300 BP suggests a similar situation may be occurring in Kosrae, especially given the similarity of the oral history narratives. More study is required, including ancient DNA sampling and a more detailed osteological analysis of the human remains sampled, and of the burial context, should it remain undisturbed.