Early Metal Age Settlement at the Site of Palemba, Kalumpang, Karama Valley, West Sulawesi
abstract
This article is about the Early Metal Age in Sulawesi, a little known period for this region. The research is based on the 2013 excavations undertaken at Palemba, a site rediscovered after being neglected for 80 years. A well-preserved occupation layer dominated by distinctive pottery sherds with ribbed patterns produced by carved paddle impressions is dated to the Early Metal Age (ca. cal. a.d. 300). With the sherds were imported beads, fragments of iron, fiber or cloth production tools, and a stone pavement which was cut by later placement of jar burials. One of these jars contained a flexed burial of a child, a type of burial never previously found in the Karama valley. Contemporary sites closer to the river mouth are badly disturbed, so Palemba provides important evidence for inland Karama valley occupation after the decline of early Neolithic settlements.
Keywords
settlement, jar burials, Kalumpang, Sulawesi, Early Metal Age
introduction
Well-preserved open settlements from sites dated to the Neolithic to Early Metal Age are extremely rare in the Indonesian archipelago. Their existence has only been indicated by well-polished artifacts found with the remains of metal objects and potsherds at the surfaces of some sites. Research on such settlements has been very limited, so it is difficult to date and define the characteristics and any cultural changes for these periods. This article discusses the transition between the Neolithic and the Early Metal Age in western Sulawesi based on the results of my excavations conducted there in 2013, following several excavations conducted by this author and others along the Karama River from 1933 to 2012.
The work conducted along the Karama River contributes original data for better understanding the changes that occurred there, how they related to the wider region of Island Southeast Asia, including the Philippines and Taiwan, and the involvement of Karama in regional networks. Both Neolithic and Early Metal Age assemblages in this region witnessed the introduction of tools involved with new techniques and domesticated plants and animals ca. 1600–800 b.c., as well as new burial practices and exotic materials in the early centuries a.d. Western Sulawesi is also interesting for [End Page 92] addressing the interplay between inland and coastal groups in the context of growing regional exchange during this crucial period.
The 2013 excavations reveal distinctive cultural assemblages for the Early Metal Age that have never been known before. Their distinctiveness (compared to the Neolithic) is suggested by the use of parallel carved-paddles in pottery production and the application of spindle whorls and bark beaters for producing fibers. Evidence of these production activities have been found at Palemba site together with burials in jars, metal objects, and imported glass and stone beads. All the evidence supports the conclusion that the inland people of the Karama valley were participating in wide exchange networks.
previous research along karama river
The first archaeological investigation of the Karama valley was conducted by P. V. van Stein Callenfels in 1933 following the report of an Amaravati-style bronze Buddha statuette that was being kept in Sampaga, a village located at the mouth of Karama River in Sulawesi (van Stein Callenfels 1951). Van Stein Callenfels excavated three archaeological sites, including a site at Sikendeng village near Sampaga and two others (i.e., Kamassi and Palemba) located 95 kminland in the vicinity of Kalumpang. Another site called Minanga Sipakko, located further downriver from Kamassi, was visited by H. R. van Heekeren during his excavation period at Kamassi in 1949. Both researchers reported the remains of Neolithic settlements dominated by plain and decorated pottery sherds and well-polished stone tools (van Heekeren 1972; van Stein Callenfels 1951).
Intensive excavations at Minanga Sipakko by a team of collaborators from Australia National University (ANU) and National Research Center for Archaeology in 2004–2007 (Simanjuntak et al. 2008) and another excavation at Kamassi in 2007–2008 by an Indonesian team (both seasons led by Simanjuntak) demonstrated the existence of Neolithic sites in the area (Anggraeni et al. 2014). Well-preserved Neolithic material culture, including red-slipped pottery, in Minanga Sipakko and Kamassi dates to ca. 1500–800 b.c. and reflects possible sources in the Philippines and Taiwan. No indication of pre-Neolithic materials has been found in these sites.
Sporadic investigations between 2004 and 2017 found new open settlement sites along the river valley. These sites, situated closer to Makassar Strait, have been disturbed by current activities and remain undated, although their archaeological assemblages are from the Early Metal Age. These sites include Sikendeng, Lattibung, the excavated Pantaraan 1 site (Anggraeni 2012; Anggraeni et al. 2014), the excavated Lemolemo site (Kirana and Darojah 2015), and the Alongalong site excavated in 2016 by Anggraeni and Hakim.
Apart from the Neolithic sites of Minanga Sipakko and Kamassi, my excavation at Bukit Pantaraan 1, Kecamatan Tommo, as part of my doctoral research program in 2008, uncovered a middle Neolithic assemblage dominated by plain unslipped pottery (Anggraeni 2012). The site was first occupied ca. 1200 b.c., and later in the Early Metal Age was reused for jar burials, as indicated by the recovery of fragments of large earthenware jars, glass beads and bracelets, and fragments of metal. Iron slag and barkcloth beaters were found on the surface of this site, but not in the excavation trenches. A glass bead and an earthenware storage jar were also reported in the upper layer of Minanga Sipakko, not in association with the Neolithic assemblages below (Simanjuntak et al. 2004). The two new sites of Sakkarra and Popanga on the [End Page 93]
Fig 1.
Map of Neolithic and Early Metal Age sites from the Karama and Bone Hau valleys in West Sulawesi (based on Ratman and Atmawinata 1993).
Bone Hau tributary of the main Karama River and one site of Salu Makulak at Kalumpang also have Early Metal Age assemblages (Mahmud et al. 2020; Suryatman et al. 2018; Tim Kajian 2014) (Fig. 1).
Palemba is another Kalumpang site that P.V. van Stein Callenfels (1951) suggested was occupied later than Kamassi based on the pottery assemblage he excavated there in 1933. Palemba was thereafter neglected by archaeologists until I visited it in 2009 and excavated it in 2013. I found a number of distinctive surface finds on the Palemba hilltop and around its western base in 2009, including a small number of decorated earthenware rims and body sherds (Anggraeni 2012). The decorated sherds were found together with pedestals, spouts, and thick plain sherds. The excavation in 2013 was undertaken to answer two main questions: How does the Palemba assemblage relate to those from the older sites of Kamassi and Minanga Sipakko? What is the importance of Palemba for explaining the development of open settlements in the Karama valley?
current research at palemba
The Palemba site is situated at 2°28′22″S and 119°29′57″E, about 100 m above sea level. Located on the south bank of Karama River, about 2 km upstream from Kamassi, it was first visited by A. A. Cense in 1933. Cense found some potsherds and was informed of the former existence of half of a big stone jar, stated to be 100 cm diameter and 50 cm high, but nothing of this jar remained at the time of his visit. His finds were reported to van Stein Callenfels (1951:90) while the latter was excavating Kamassi in 1933.
Van Stein Callenfels excavated Palemba the same year, following his initial excavation at Kamassi. His excavation at Palemba revealed lots of plain sherds, but the [End Page 94] decorated (including red-slipped) sherds and stone implements that he had found in abundance at Kamassi and Minanga Sipakko were absent from Palemba (van Stein Callenfels 1951). Van Stein Callenfels suggested that an earthenware figurine with a human head found at Palemba was comparable in design and date to an earthenware sherd with a human face in relief from Kamassi. These sherds of earthenware with human faces were thought to be younger than the other decorated pottery in Kamassi (van Stein Callenfels 1951:90) and it was even suggested that the plain sherds found at Palemba were similar to modern Toraja pottery.
In contrast to van Stein Callenfels' finds, my site visit in January 2009 yielded a small number of surface finds including pottery rims with incised and punctate decoration. I have suggested previously that these were comparable to rims from the second and final phases of both the Kamassi and Minanga Sipakko Neolithic sites. However, sherds of thick pottery and a small number of body sherds with bi-directional ribbed decoration appear to be Early Metal Age (Anggraeni 2012) (Fig. 2).
Fig 2.
Rim profiles of pottery found on the surface at Palemba (after Anggraeni 2012: fig. V.58).
[End Page 95]
In June 2013, I and my team conducted a surface survey at Palemba to locate the most promising location for excavation. The site covers an area of about 1.39 ha at the confluence of the Karama and Bone Karataun rivers (Fig. 3). The finds come from a hill covered by Imperata cylindrica and a flat area below the hill is used for cultivation and a cacao plantation. The site has some planted coconut and teak trees, while paper mulberry (Broussonetia papyrifera or Morus papyrifera L.) and Canarium sp. grow wild.
Most of the surface finds were discovered along an artificial embankment in the cacao plantation, at the western base of the hill, and on the hill slope. Sherds, mostly plain and more than 1 cm thick, were found together with a fragment of a bark-cloth beater with fine parallel incisions on both surfaces. A number of distinctive traditional burials occur in a cultivated area west of the hill; they are marked by rectangular stone formations oriented east-west with a single stone erected at one end. This place is called Palemba, meaning 'graveyard' in the Kalumpang dialect, but the Tobarak (traditional leaders) of Kalumpang do not have any information about the people buried there. The stone formations are different from Christian burials (most of the Kalumpang people are Christian), but their east-west orientation is comparable to burial practices of the Makassar people in the fourteenth century (Bulbeck 2004:224). The stone formations also remind me of similar ones from Lahat in South Sumatra that mark the burials of village founders.
The results of the surface survey convinced me to open an excavation area in a relatively flat part of the western foot of the Palemba hill. The excavation covered 5 m2 and consisted of squares aXXV, AX, aVIII, n5, and k17, which were excavated to
Fig 3.
Topographic map of Kalumpang and Tambingtambing (2013-52, 2013-61) showing Palemba site and area excavated in 2013 (modified after Bakosurtanal 1991).
[End Page 96]
Table 1.
Distribution of Pottery Sherds and Other Finds from Square aXXV, Palemba
[End Page 97] 210 cm maximum depth. Each 1 × 1 m2 square was excavated by spits. The first from the surface was 30 cm thick, and subsequent ones were 10 cm. The sediments were both dry and wet-sieved.
The stratigraphy comprises three layers: a thin topsoil (Layer 1), a 100 cm thick occupation and cultural layer (Layer 2), and a sandy basal layer (Layer 3). Two features were discovered in the well-preserved cultural layer, these being a stone pavement and two jar burials. The occupation layer surrounded the stone pavement in Square aXXV, with the top 20 cm being slightly disturbed (Table 1). The assemblage was dominated by earthenware sherds, either plain or decorated with ribbed impressions made with a carved paddle.
The stone pavement, with a north-south orientation, started to appear at 60 cm below the surface. It consisted of regular-sized rounded river boulders, 20 cm diameter on average. The pavement was about 90 cm wide but of unknown length; it seems to extend further north and south than what was excavated (Fig. 4a). It could be a house foundation or part of other constructions in a settlement (Bellwood 2017:248, plate 5). There were no other finds except those mentioned below nor indications of burials underneath. Further investigation is needed to reveal the whole formation and explain its function.
Most archaeological finds other than sherds were distributed at 50–90 cm depth, immediately above the stone pavement or among and below the boulders. They comprise a carnelian bead, 11 monochrome glass beads, a fragment of iron, an
Fig 4.
Findings from Square aXXV at Palemba: (a) Boulder pavement 60–80 cm below surface and associated artifacts; (b) monochrome glass beads; (c) carnelian bead; (d) paddle impressed sherd; (e) halfcircle pottery handle; (f) bark-cloth beater; (g) iron fragment; (h) fossil wood object (photos b–c, g–h by Sheila Ayu Rachmadiena; photos a, d–f by Anggraeni).
[End Page 98] unidentified rectangular semi-fossilized wooden object, a very small flake of silicified limestone, and two fragments of a bark-cloth beater found less than 10 cm apart (Fig. 4). Another glass bead was found at about 100 cm depth. No stone axes or flaked tools were found during the excavation. The semi-fossilized wooden artifact (Moh's scale hardness 3) is similar in size and general shape to stone chisels found at Minanga Sipakko and Kamassi, but has no sharp working edge.
The cultural layer was dated to 1720 ±30 b.p. (Beta-360430), calibrated to a.d. 248–391 (cal. 7.1 at two sigma, 95.4% probability range), from carbonized residue on a sherd recovered 90 cm below the surface in Square aXXV. The laboratory report indicates this as a minimum age (pers. comm. Chris Patrick, Beta Analytic 2013). No other samples can be used for dating analysis, including the human remains from the jar burials.
The two burial jars were uncovered 14 m to the east of the stone pavement. The first was an undisturbed burial in a big jar that started to appear 70 cm below the surface in Square AX. The second jar was found broken, only 1m away from the first, but about 100 cm below the surface. Fragments of human bone and teeth found outside and above the broken jar in Square aVIII indicate past disturbance. The jars were associated with a continuing occupation layer with many plain and paddle impressed sherds like those elsewhere in the site, more glass beads, and a bark-cloth beater. A fragment of a terracotta spindle whorl was also found in the occupation layer about 60 cm below the surface. This is the only one ever found in the Karama sites.
The Jar Burials
The first burial jar was well-fired, about 70 cm in maximum diameter and 40 cm rim diameter, with sides a minimum of 4 mm thick. It had been covered with a flat stone 12 cm thick and 40 cm in diameter, together with three other stones and some gravel. The heavy stones placed on top caused the jar rim and shoulder to collapse, leaving about three quarters of the height still standing to 37 cm (Fig. 5a). Diagonal incised lines and punctate impressions were applied around the flat lip of the jar rim (Fig. 5b).
The jar had apparently been deliberately chiseled horizontally right around the shoulder for insertion of the burial, after which the two parts of the jar were fitted back together again (Bellwood and Dizon 2013:62–64). It was supported upright in a hole dug through the occupation layer with stones placed around its rounded base. A similar technique was used for standing the second jar. The burial jar at Pantaraan 1 was also placed in a hole dug into the basal natural gravel (Anggraeni 2012).
A flexed skeleton of a child was found intact in the bottom of the first burial jar (Fig. 5c). The skeletal remains were fragmented and very poorly preserved. Adisuriyanto (pers. comm. 2013) suggests that the child was about 4 to 5 years old. A dark blue glass bead was found at the child's wrist; presumably it was originally tied on to the wrist (Fig. 5d). The bead may have been an amulet to protect the child from childhood evil. This was until very recently a common practice among many ethnic groups in Indonesia, including the Dayaks, Buginese, and Javanese people.
The second jar retained less than half of its base and wall. Five human teeth (incisor, canine, premolar, second and third molars), all without roots and perhaps belonging to an adult, were found in and around it. Two glass beads were also in the jar, with seven scattered in the adjacent occupation layer. The absence of human bones from the jar burial evokes the burial practice of interring disarticulated teeth, which was common [End Page 99]
Fig 5.
Intact jar burial from Square AX at Palemba: (a) large flat stone lid; (b) decorated flat jar lip; (c) flexed child burial in the base; (d) dark blue glass bead found near child's wrist (photos by Anggraeni).
among people of Island Southeast Asia in the late prehistoric period (Bulbeck 2004:224).
The teeth with this second burial jar indicate that the person buried was about 20–24 years old, based on the degree of second and third molar dental attrition. Mild caries were noted on the third molar. The shovel-shaped incisor cross-sections indicate that the owner had Asian rather than Australo-Papuan cranio-dental affinities. The third molar also had stains from areca nut chewing.
The previous jar burials excavated in the Karama valley from the site of Pantaraan 1 were badly disturbed and without bones or teeth. Unfortunately, the Palemba child burial retained no collagen and so could not be dated. This means that we have no 14C dates yet for any Karama valley jar burials.
The Palemba Pottery
Besides the burial jars, 3812 sherds were found during the 2013 Palemba excavation. They were distributed from just below the surface down to 120 cm in depth. My examination of the earthenware sherds suggests two functional vessel groups: [End Page 100]
Fig 6.
Distribution of decorated and plain sherds at Palemba.
containers and non-containers. Container-related sherds consist of plain and decorated rims, plain and decorated body sherds, 2 lids, 18 handles, 6 necks, 1 spout, and 4 bases or feet (Fig. 6). Non-container earthenware items comprise 6 stove fragments and a spindle whorl (Table 2).
The pottery from Palemba is dominated by restricted rims with internally convex and thickened lips. Most are short, thickened, and rounded (see Fig. 2, rims B1–B6). Comparable rims occur in the late Neolithic assemblages at Kamassi and Minanga Sipakko and at Pantaraan 1. Flattened lips occur only on a few open bowls and the complete burial jar.
As in the older Kamassi and Minanga Sipakko Neolithic assemblages, decorated pottery was rare at Palemba. Most vessel bodies were unslipped and plain, but 7.1 percent of all body sherds were impressed with a carved paddle. Parallel grooves were impressed horizontally and vertically by these paddles to create patterns that did not exist in the Neolithic assemblages (Simanjuntak et al. 2008) (Fig. 4c). This decoration was applied directly below the rim and over the rest of the body. However, open bowls with carinations were decorated by parallel diagonal incised lines to form triangles above the carinations (Fig. 2 rims F3–F4, Fig. 7a).
As in the upper layers at Minanga Sipakko and Kamassi, the lip decoration techniques were quite varied at Palemba. They include punctation, notching, incision, and finger impression (Fig. 7b). Perforations sometimes occur below lips or through pedestals. The intact burial jar has incision combined with punctation on its flat lip. Criss-cross incised decoration also occurs on some flat lips (Fig. 7c), showing a parallel with the Late Neolithic assemblage at Kamassi. [End Page 101]
Table 2.
Distribution of Sherds from all Excavated Squares at Palemba
[End Page 102]
Fig 7.
Decorated rim sherds from Palemba: (a) incised and notched; (b) finger impressed; (c) criss-cross incision (photos by Anggraeni).
The handles of lids or container vessels from Palemba are less diverse than those from Minanga Sipakko and Kamassi. Semi-circular horizontal or vertical handles are common and resemble handles from Pantaraan 1. Other Early Metal Age sites that have recently been discovered downstream from Kalumpang, such as Popanga (Tim Kajian 2014) and Alongalong and Lemolemo (discovered by Anggraeni and Budianto Hakim in 2016) also include this type of handle. The Nagsabaran shell midden in northern Luzon also has similarly-shaped handles on unrestricted bowls (Hung 2008).
I and my former student, Riana Wulan Pradipta, analyzed the chemical components in the pottery clay used at Palemba with a Scanning Electron Microscope (SEM) at Universitas Gadjah Mada (UGM). Seven elements (Mg, Al, Si, K, Ca, Ti, Fe) were chosen for the analysis. Comparative samples were analyzed from the Palemba and Pantaraan 1 burial jars and from other Early Metal Age sites in the Karama valley (i.e., Lemolemo 3, Alongalong, Sakkarra, Popanga). A potential clay source sample was taken from the Salu Taranusi stream bed. All results were analyzed statistically using the Canonical Discriminant Function method.
The results indicate three possible groups. Except for the Pantaraan 1 jar and one sherd from Popanga, all samples belong to one local geochemical group related to the Salu Taranusi clay sample. The other two sherds are both unique in having higher percentages of magnesium, and the Pantaraan 1 jar has lower aluminium oxide. All the pottery was tempered with volcanic sand. The overall results support my previous suggestion that the Karama pottery was produced locally (Anggraeni 2012).
Fiber and Cloth Production
An important find among the baked clay specimens found at Palemba is a fragment of a decorated spindle whorl, an object never reported before from the Karama valley, or indeed Sulawesi. It has a plano-convex cross section, with a slightly elevated rim around its flat side, and a perforation in the center. Punctate impressions and grooves occur around the rim (Fig. 8c). The specimen is 6 cm in diameter, with a 1.4 cm diameter central perforation, 1.32 cm in maximum thickness, and 30 grams in weight. Its shape fits Type II in Judith Cameron's (2011:562, fig. 3) Southeast Asian spindle whorl typology.
Cameron (2011) suggested that spindle whorl weights and sizes correlate with the characteristics of the fibers. Spindle whorls are components of hand spinning where the whorls keep the spindles (shafts) vertical when they rotate. Heavy whorls (more than 15 g) are needed to spin strong, coarse fibers (Cameron 2011:563). Based on its [End Page 103]
Fig 8.
A fragment of a spindle whorl: (a) convex surface; (b) flat surface with elevated band; (c) detail of elevated band with incised and punctate decoration.
estimated weight, the whorl from Palemba could be used for spinning fibers derived from inner tree bark, as has been recorded in a number of ethnographic reports.
Ethnographic records as well as archaeological research show that various kinds of tools such as spindle whorls or bark-cloth beaters play important roles in producing textiles from different fibers (Cameron 2011; Oley 2007). Bark beaters can be used for separating fibers of the inner tree bark to prepare fabrics. However, producing cloth using spindle whorls is a longer process than beating the bark into fabric, since the pounded fibers then need to be spun and weaved. In Borneo, ethnographic studies have shown some groups still use the inner bark from available trees and shrubs to produce woven textiles (Oley 2007). Paper mulberry (Broussonetia papyrifera) has played an important role in the production of bark cloth that has long been worn in Sulawesi, while minority groups in South China and a small number of Han Chinese have used the bast (inner bark) of paper mulberry and other fiber-producing species to make woven textiles. The fibers were first boiled in lime to soften them and spliced (Kuhn 1988:56–57). The use of paper mulberry for both kinds of textile production was also found among the Banaue people from Ifugao Province in the Philippines (Milgram 2007).
Both bark cloth production and spinning first occur in South China around 7000 years ago and spinning and weaving appear to have developed alongside bark-cloth technology. Beyond Taiwan, baked clay spindle whorls were found in a number of Neolithic and Early Metal Age sites in Island Southeast Asia (Cameron 2001, 2013), although they have rarely appeared with bark-cloth beaters. One of the exceptions was Arku Cave in northern Luzon (2200–50 b.c.), where a bark-cloth beater was found together with spindle whorls and other artifacts as burial gifts (Thiel 1988–1989:78).
It is interesting that two bark-cloth beaters are found in the same occupation layer at Palemba site. In contrast to spindle whorls, bark-cloth beaters have been found at almost all prehistoric sites in the Karama and Bone Hau valleys dated to the Late Neolithic and Early Metal Age (Anggraeni 2012; Fakhri et al. 2015:20; Suryatman et al. 2018:209). Except for those from Minanga Sipakko, none have handles. All beaters have parallel grooves on both functional surfaces (see Fig. 4f). The upper and lateral sides of the beaters without handles have grooves around them for hafting. Such beaters are still used today for preparing Ficus sp. bark cloth in central Sulawesi.
Some older Karama villagers can still recall the bark cloth worn by the older people in Tamemongga village, close to the Pantaraan site, and in the upstream villages of Karataun and Tambingtambing (Anggraeni 2012). However, spindle whorls are no [End Page 104] longer used. One of the elders, an eighty-year-old man named Sanusi (pers. comm. November 2008), mentioned that sekomandi, the famous handwoven ikat fabrics still made by women of Karataun and Kalumpang, were originally made of inner tree bark fiber. Cotton has only been known very recently in those areas.
The existence of a terracotta whorl together with bark-cloth beaters in the occupation layer at Palemba confirms Cameron's (2001) suggestion that the two artifacts had different functions in Island Southeast Asia, especially during the Early Metal Age. Cameron (2001) states that woven textiles had an important role in the social life of early Austronesian communities, as they still do today. Spindle whorls dating from the Metal Age in Island Southeast Asia may also have been used to spin other fibers widely used for textiles at this time, including ramie (Boehmeria nivea), hemp (Cannabis sativa), and cotton (Gossypium sp.) (Cameron pers. comm. 2020).
Based on the wide spread of spindle whorls during the Neolithic and Early Metal Age sites in Southeast Asia and the existence of a terracotta whorl at Palemba, further research on textile productions during the prehistoric period in the Karama sites should be conducted in future. It is likely that perforated stone artifacts ca. 1.5–2cm in diameter were originally used as spindle whorls in Neolithic contexts at Minanga Sipakko and Kamassi, comparable to Types I and VIII of Cameron's Southeast Asian whorl typology. Examples of such potential stone whorls were found by Truman Simanjuntak and his team at Minanga Sipakko in 2004 (Anggraeni 2012) and by Budianto Hakim in the 2011–2012 excavation at Kamassi (Hakim 2014; Mahmud et al. 2020:81).
Subsistence at Palemba
The 2013 excavation at Palemba yielded only a very small number of animal bones and teeth, identified as pig, deer, monkey, and fowl (possibly chicken), in the occupation layer in Square aVIII. The same species were found in both the Minanga Sipakko and Kamassi Neolithic sites, with pigs being the most abundant and perhaps managed (Anggraeni 2012; Anggraeni et al. 2014).
Dental calculus from the child buried in the first Palemba jar was examined by Rahayuningsih (2014), who identified starch grains, including taro, arrowroot, and breadfruit. As at Kamassi (Anggraeni 2012; Anggraeni et al. 2014) and Minanga Sipakko (Anggraeni et al. 2012), I and my student, R. Linda Octina, extracted phytoliths from Palemba; we identified a small number of fan-shaped (bulliform) rice phytoliths in the occupation layer at 60 cm below the surface (Fig. 9). The small number of phytoliths from rice leaves may indicate some possibilities. First, rice fields in the Kalumpang region might be located far from the settlement areas, as now. Traditional farmers cut panicles during harvesting and left the stalks and leaves in the rice field, while threshing and winnowing were carried out in both the field and settlement areas. It is not therefore surprising that very few rice phytoliths were found in the settlement area at Palemba. Second, the stratigraphic layers show frequent inundation by the Karama River. Most microscopic remains of plants, including rice phytoliths might have been washed out by flood. Third, rice might be less important at Palemba than tubers and other staple foods identified from the dental calculus. [End Page 105]
Fig 9.
Fan-shaped rice phytoliths from Palemba (scale in microns) (image taken by Octina).
discussion: palemba in regional context
Current research along the Karama River provide archaeological associations useful for recognizing the development of prehistoric settlements in the river valley. The Karama River itself is about 150 km long. It served in the past, as now, as an important arterial route for human interaction in West Sulawesi. Until very recently, people of inland regions such as Kalumpang brought cocoa, rice, other staple vegetable foods and rattan and other plants harvested from the forest down the river to sell in the coastal region of Makassar Strait.
The results of current research suggests that upstream Neolithic settlements were settled first by rice farmers, followed by the establishment of Early Metal Age settlements downstream as sea-level stability and alluviation produced more land suitable for cultivation (Anggraeni 2012; Deng et al. 2020; Fakhri et al. 2015; Suryatman et al. 2018). In the upstream Karama valley, the only Metal Age site known so far is Palemba, which shows fairly sparse occupation remains.
Palemba has an important role to play in explaining the transition between Neolithic and the Early Metal Age periods, not only in the context of Karama valley, but also for Indonesia in general. Most Early Metal Age sites were found disturbed and overlapped with more recent activities. It is difficult to distinguish artifacts (especially pottery) between Neolithic and Early Metal Age sites when they have been disturbed. However, the changes in techniques of pottery decoration and rim cross-section from the Early Neolithic to the Early Metal Age are obvious at Palemba. Carved paddle impression, also known as "basket-" or "paddle-marked" by Hung (2008:162–171), occurs in Neolithic contexts at Nagsabaran in the Cagayan Valley (Hung 2008), but is so far absent in the Karama Valley Neolithic sites of Kamassi and Minanga Sipakko. Such carved paddle impression pottery dominates the Metal Age contexts at Palemba and other Sulawesi sites such as Popanga at the confluence of the Bone Hau and Karama rivers (Tim Kajian 2014) and the Gua Mo'o hono shelter in the Towuti-Routa region of southeastern Sulawesi (O'Connor et al. 2018:128). In Borneo, this kind of pottery decoration is dominant in Metal Age contexts at Niah West Mouth, Gua Sireh, Madai Cave, and Nanga Balang (Bellwood 2017:325, fig. 9.8(d); Wibisono 2006).
The Island Southeast Asian jar burial tradition is first reported from the Middle Neolithic of Taiwan at about 2000 b.c.e., then becomes more common during the Late [End Page 106] Neolithic ca. 1000 b.c.e. at sites such as Huagangshan on the eastern coast of Taiwan (Hung 2017:238). As well as at Palemba, other Metal Age jar burial traditions occur in Vietnam, Laos, South Sumatra, the Philippines, Borneo, Java, Bali, and Sumba (Bellwood 2017; Bulbeck 2017; Indrawooth 1997). According to Hung (2017:238), Metal Age burial jars in the Batanes Islands in the Philippines were "identical to many in eastern Taiwan."
Bulbeck (2017:154) has summarized three "horizons" or phases of jar burial in the Indo-Malaysian region. The Palemba type, big mortuary jars with stone covers, seems to belong to his second horizon, dated to the early centuries a.d. (Bellwood 2017; Bulbeck 2017:154). This fits well with the 14C date from the Palemba site. Primary burials in big jars with fairly similar dates can also be found in Batanes, Mindanao, Talaud, Maluku, Sumba, and Sumatra (Bulbeck 2017).
The bark-cloth beaters with parallel grooves on both surfaces from Palemba and other sites in the Karama and Bone Hau valleys are comparable to examples of Type IV beaters from the Asian mainland and South China (Cameron 2008:205, fig. 3). The decorated terracotta spindle whorl from Palemba is also comparable to specimens from South China and Mainland Southeast Asia (Cameron 2011:564). The co-existence of a spindle whorl and a bark-cloth beater in the same layer at Palemba emphasizes that two different tools of cloth production were in use.
Exotic artifacts include monochrome glass and carnelian beads and metal objects. These indicate links to an Asian trading network during the early centuries a.d., possibly down the Karama valley. Carnelian beads are very rare elsewhere in the Karama valley, with one found on the surface at Latibung and one in a wooden coffin in Tabuun Cave. Current research suggests that carnelian beads were not only produced in South Asia, but in Southeast Asia as well, where they are known to have existed in the last centuries b.c. (Bellina 2003). Examinations of quality, size, and perforations on single shape carnelian beads found in the Karama valley imply that they can be included in the second period of Southeast Asian production, dating to the early millennium a.d. (Bellina 2003:289–291).
The production and distribution of glass beads in South and Southeast Asia were more complex than what has previously been thought (Calo et al. 2015; Carter 2016). Unlike the white glass beads from a number of African and South Asian sites mentioned by Dussubieux and colleagues (2008) and Carter (2016), Palemba beads were not deliberately white. Some opaque red glass beads seem to have been lower quality than the others, turning whitish when washed. White glass beads, identified as Asian potash glass, also appeared among the cultural assemblages of the Pacung burial phase in the north coast of Bali (Calo et al. 2015), but there is no further evidence for the production of white beads in the region.
conclusion
The cultural remains from Palemba reveal separate cultural influences from those present during the Karama valley Neolithic. The Early Metal Age settlement at Palemba likely occurred ca. 240–400 a.d. and quite possibly predated the big jar burials. The evidence is ambiguous because the jars were put in place by digging holes that cut through the occupation layer.
The existence of the carved paddle impression technique at Palemba puts the site into a wider regional context. The technique could have been introduced by different [End Page 107] groups from Taiwan, Nagsabaran, or other places in Southeast Asia. It seems that the technique entered the Karama valley together with the introduction of iron objects. More attention should be paid to the use of baked clay spindle whorls in tandem with bark-cloth beaters in the Early Metal Age of the Karama valley. Remains of the whorls might occur among the sherds found at other sites.
Palemba and Sakkarra in upstream Karama, along with downstream Metal Age sites such as Latibung, Lemolemo, and Pantaraan 1, were indirectly linked to regional trade networks involving carnelian and glass beads during the early centuries a.d. that further linked them to Mainland Southeast Asia and possibly southern China and South Asia. A number of questions should be addressed. What roles did the coastal and inland groups play in these exchange networks? Did the inland people provide rice and other staples, plant and forest products, or iron ore and gold to coastal peoples in the past, as has been demonstrated until today? Iron ore is still mined along the Bulo River at Kondo Bulo village and the traditional gold mine located at Batuisi, 8.8 km east of Kalumpang, is still in use today (Anggraeni 2012). Did the coastal people offer useful products such as salt, dried fish, or sarongs to their inland counterparts in the past as they still do today? Did they also manipulate the rules and take advantage of their trade partners in the networks to fulfill the demands for imported goods? Further investigations are needed in future to understand the nature of exchange networks during the Early Metal Age in this region.
Anggraeni
Anggraeni is an Assistant Professor at the Department of Archaeology, Faculty of Cultural Sciences, Universitas Gadjah Mada in Yogyakarta.
acknowledgments
My special thanks go to the Faculty of Cultural Sciences, Gadjah Mada University, for providing funding for the research at Palemba in 2013; to my excavation team, Restu Ambar Rahayuningsih, Eko Setyo Pranoto, and Arundina Ardhanari; to Darius and his sons from Kalumpang, as my local research team; to the Tobarak of Lembang Kalumpang, Eli Sipayo, who provided information about current Kalumpang inhabitants and culture; to R. Linda Octina, who help me prepare samples for phytolith analysis; to Martha Luhana Herbiamami, Rusyad Adi Suriyanto, Ashwin Prayudi, and Novialita Putri Ridimas, who helped with bone analysis at UGM; to staff at the UGM Integrated Laboratory (LPPT) who allowed me to use SEM equipment for pottery analysis; and to Riana Wulan Pradipta, who helped me analyze pottery. Thanks also to Peter Bellwood for his valuable comments on a draft of this article and Judith Cameron for her suggestion concerning ancient textile production. I presented results of the Palemba research at the IPPA Congress in Siem Reap, Cambodia, 12–18 January 2014.
references cited
Anggraeni 2012 The Austronesian Migration Hypothesis as seen from Prehistoric Settlements on The Karama River, Mamuju, West Sulawesi. Ph.D. thesis. The Australian National University, Canberra.
Anggraeni, Muasomah, R. Octina, and G. Sukma 2012 Variasi vegetasi di pemukiman prasejarah Minanga Sipakko, Sulawesi Barat, berdasarkan identifikasi fitolit [Vegetation Variation in Prehistoric Settlement of Minanga Sipakko, West Sulawesi, based on Phytolith Identification]. Unpublished report for Department of Archaeology, Gadjah Mada University, Yogyakarta.
Anggraeni, Truman Simanjuntak, Peter Bellwood, and Philip Piper 2014 Neolithic foundations in the Karama Valley, West Sulawesi, Indonesia. Antiquity 88:740–756.
Bakosurtanal 1991 Sheet 2013-52 (and) Sheet 2013-61, in Topographic Maps of Kalumpang and Tambingtambing. Bogor: Badan Koordinasi Survei dan Pemetaan Nasional (Bakosurtanal).
Bellina, Bérénice 2003 Beads, social change and interaction between India and South-east Asia. Antiquity 77:285–297, doi:10.1017/S0003598X00092279.
Bellwood, Peter 2017 First Islanders. Hoboken, NJ: John Wiley and Son.
Bellwood, Peter, and Eusebio Dizon 2013 Archaeological investigations at Savidug, Sabtang Island, in 4000 Years of Migration and Cultural Exchange: The Archaeology of the Batanes Islands, Northern Philippines:47–65, ed. Peter Bellwood and Eusebio Dizon. Terra Australis 40. Acton, ACT: ANU-E Press.
Bulbeck, David 2004 South Sulawesi in the corridor of island populations along East Asia's Pacific rim, in Quaternary Research in Indonesia: 221–258, ed. S. G. Keates and J. M. Pasveer. Leiden: A.A. Balkema.
2017 Traditions of jars as mortuary containers in the Indo-Malaysian Archipelago, in New Perspectives in Southeast Asian and Pacific Prehistory: 141–164, ed. Philip J. Piper, Hirofumi Matsumura, and David Bulbeck. Terra Australis 45. Acton, ACT: ANU Press.
Calo, Ambra, Bagyo Prasetyo, Peter Bellwood, James W. Lankton, Bernard Gratuze, Thomas Oliver Pryce, Andreas Reinecke, Verena Leusch, Heidrun Schenk, Rachel Wood, Rochtri A. Bawono, I Dewa Kompiang Gede, Ni L. K. Citha Yuliati, Jack Fenner, Christian Reepmeyer, Cristina Castillo, and Alison K. Carter 2015 Sembiran and Pacung on the north coast of Bali: A strategic crossroads for early trans-Asiatic exchange. Antiquity 89:378–396.
Cameron, Judith 2001 Textile technology and Austronesian dispersals, in The Archaeology of Lapita Dispersal in Oceania: Papers from the Fourth Lapita Conference: 177–183, ed. G. R. Clark, A. J. Anderson, and T. Vunidilo. Terra Australis 17. Canberra: Research School of Pacific and Asian Studies
2008 Trans-oceanic transfer of bark-cloth technology from South China–Southeast Asia to Mesoamerica?, in Islands of Inquiry Colonisation, Seafaring and the Archaeology of Maritime Landscapes: 203–210, ed. Sue O'Connor, Geoffrey Clark, and Foss Leach. Terra Australis 29. Canberra: Research School of Pacific and Asian Studies
2011 Iron and cloth across the Bay of Bengal: New data from Tha Kae, Central Thailand. Antiquity 85:559–567
2013 The spinning tools from Sunget, Anaro, and Savidug, in 4000 Years of Migration and Cultural Exchange: The Archaeology of the Batanes Islands, Northern Philippines: 115–112, ed. Peter Bellwood and Eusebio Duzon. Terra Australis 40. Acton ACT: ANU-E Press.
Carter, A. K. 2016 The production and exchange of glass and stone beads in southeast Asia from 500 BCE to the early second millennium CE: An assessment of the work of Peter Francis in light of recent research. Archaeological Research in Asia 6:16–29, dx.doi.org/10.1016/j.ara.2016.02.004.
Deng, Zhenhua, Hsiao-chun Hung, Mike T. Carson, Adhi Agus Oktaviana, Budianto Hakim, and Truman Simanjuntak 2020 Validating earliest rice farming in the Indonesian Archipelago. Scientific Reports 10:10984, doi:10.1038/s41598-020-67747-3.
Dussubieux, L., C. M. Kusimba, V. Gogte, S. B. Kusimba, B. Gratuze, and R. Oka 2008 The trading of ancient glass beads: New analytical data from South Asian and East African soda–alumina glass beads. Archaeometry 50(5):797–821.
Fakhri, Suryatman, Budianto Hakim, and Ratno Sardi 2015 Exploration of prehistoric sites in the Karama Watershed, West Sulawesi, Indonesia: From early occupation until the metal age. Journal of Indo-Pacific Archaeology 39:18–25.
Hakim, Budianto 2014 Archaeological traces of Austronesian ancestors at the Kamansi site of the Karama River Valley in West Sulawesi, Indonesia. Journal of Austronesian Studies 5(1):73–95.
Hung, Hsiao-chun 2008 Migration and Cultural Interaction in Southern Coastal China, Taiwan and the Northern Philippines, 3000 B.C. to A.D. 100: The Early History of the Austronesian-Speaking Populations. Ph.D. thesis. Australian National University, Canberra.
2017 Neolithic cultures in southeast China, Taiwan, and Luzon, in First Islanders: 232–240, ed. Peter Bellwood. Hoboken, NJ: John Wiley and Sons.
Indrawooth, Phasook 1997 The practice of jar burial in the Mun and Chi Valleys. Bulletin of Indo-Pacific Prehistory Association: Chiang May Papers 3:149–152.
Kirana, Ayu Dipta, and Citra Iqliyah Darojah 2015 Report of Research in Austronesian Neolithic Site in Karama River Valley. Unpublished report to Granucci Fund, archived at the Australian National University, Canberra.
Kuhn, Dieter 1988 Textile Technology: Spinning and Reeling. Science and Civilization in China (series), vol. 5, part 9: Chemistry and Chemical Technology, ed. Joseph Needham. Cambridge: Cambridge University Press.
Mahmud, M. Irfan, Budianto Hakim, Fakhri, Ratno Sardi M., Suryatman, and Andi Muhammad Saiful 2020 Kebudayaan Kalumpang Sulawesi Barat (3,800 B.P.–400 A.D.) [Kalumpang Culture, West Sulawesi (3,800 B.P.–400 A.D.)]. Makassar: Balai Arkeologi Sulawesi Selatan.
Milgram, B. 2007 Re-crafting tradition and livelihood: Women and bast fibre textiles in the upland Philippines, in Material Choices: Refashioning Bast and Leaf Fibres in Asia and the Pacific: 119–132, ed. R. Hamilton and B. Milgram. Los Angeles: Fowler Museum of Cultural History, UCLA.
O'Connor, Sue, David Bulbeck, Philip J. Piper, Fadhila Aziz, Ben Marwick, Fredeliza Campos, Jack Fenner, Ken Aplin, Fakhri, Suryatman, Tim Maloney, Budianto Hakim, and Rachel Wood 2018 The human occupation record of Gua Mo'o hono shelter, Towuti-Routa region of Southeastern Sulawesi, in The Archaeology of Sulawesi: Current Research on the Pleistocene to the Historic Period: 117–151, ed. Sue O'Connor, David Bulbeck, and Juliet Meyer. Terra Australis 48. Acton, ACT: ANU Press.
Oley, Elizabeth 2007 Ulap Doyo, in Material Choices: Refashioning Bast and Leaf Fibres in Asia and the Pacific: 62–78, ed. Roy Hamilton and B. Lynne Milgram. Los Angeles: Fowler Museum of Cultural History, UCLA.
Rahayuningsih, Restu Ambar 2014 Makronutrien dalam Makanan Manusia Pendukung Situs Petawua Anditu dan Wineki, Kabupaten Poso, Sulawesi Tengah [Macronutrients in the Human Diet of the Petawua Anditu and Wineki Sites, Poso District, Central Sulawesi]. B.A. thesis. Fakultas Ilmu Budaya, Universitas Gadjah Mada, Yogyakarta.
Ratman, N., and S. Atmawinata 1993 Systematic Geological Maps, Indonesia, Sheet Mamuju - 2013.
Simanjuntak, Truman, Kerrie Grant, Fadhlan S. Intan, Nani Somba, Irfan Machmud, Bernadetta Akw, Danang Wahyu Utomo, Ngadiran, and M. J. Morwood 2004 Excavation Report 2004, Minanga Sipakko Site, South Sulawesi. Unpublished report, archived at the National Research Centre for Archaeology, Jakarta.
Simanjuntak, Truman, M. J. Morwood, Fadhlan S. Intan, Irfan Machmud, Kerrie Grant, Nani Somba, Bernadetta Akw, and Danang Wahyu Utomo 2008 Minanga Sipakko and the Neolithic of the Karama River, in Austronesian in Sulawesi: 57–75, ed. Truman Simanjuntak. Depok: Center for Prehistoric and Austronesian Studies.
Suryatman, Budianto Hakim, and Fakhri 2018 The Sakkarra site: New data on prehistoric occupation from the Metal Phase (2000 B.P.) along the Karama drainage, West Sulawesi, in The Archaeology of Sulawesi: Current Research on the Pleistocene to the Historic Period: 201–221, ed. Sue O'Connor, David Bulbeck, and Juliet Meyer. Terra Australis 48. Acton, ACT: ANU Press.
Thiel, Barbara 1988–1989 Excavations at Musang Cave, Northeast Luzon, Philippines. Asian Perspectives 28(1):61–81.
Tim Kajian 2014 Kajian Penyelamatan Situs Cagar Budaya di Daerah Aliran Sungai Karama dan Bone Hau [Rescue study of the heritage in Karama and the Bone Hau watershed]. Report for Direktorat Jenderal Kebudayaan, Direktorat Pelestarian Cagar Budaya dan Permuseuman, Kementerian Pendidikan dan Kebudayaan, Jakarta.
van Heekeren, H. R. 1972 The Stone Age of Indonesia. Hague: Nijhoff.
van Stein Callenfels, Pieter Vincent 1951 Prehistoric sites on the Karama River. University of Manila Journal of East Asiatic Studies 1(1): 82–97.
Wibisono, Sonny C. 2006 Stylochronology of early pottery in island of Southeast Asia: A reassessment of archaeological evidence of Austronesian, in Proceedings of the International Symposium: Austronesian Diaspora and the Ethnogeneses of People in Indonesian Archipelago: 107–118, ed. Truman Simanjuntak, Ingrid H. E. Pojoh, and Mohammad Hisyam. Jakarta: LIPI Press.
