2. PRELIMINARY RESULTS FROM THE 1998-99
FIELD SEASONS IN LUWU BY THE OXIS PROJECT
(ORIGINS OF COMPLEX SOCIETY IN SOUTH SULAWESI)
F. David Bulbeck, Dept. of Archaeology and Anthropology, Australian National University and Bagyo Prasetyo, Bidang Prasejarah, Pusat Penelitian Arkeologi Nasional

Page 3

(3) Three main advantages of the Malangke coastal region underwrote Luwu's establishment and efflorescence here,

1. its agricultural potential for producing sago to support a large population,

2. its strategic location at the mouth of the Baebunta River which could be used for shipping highlands produce-high-grade iron ore, gold and dammar- transported overland a short distance from the Rongkong River within the area controlled by the Baebunta chiefdom,

3. the production of iron weapons and agricultural tools in Pattimang.

Expectations: Sago phytoliths will be abundant in 14th century and later contexts at Malangke coastal sites. Excavations in Pattimang and Baebunta will reveal traces of iron which can be sourced to the upper Rongkong, as well as gold and dammar gum. Numerous iron artifacts, as well as a factory area for producing ironwares, will be datable to at least the 15th century at Pattimang.

Evidence: The hypothesized advantages of coastal Malangke are confirmed to some degree. Although phytolith identifications have not yet been undertaken, the sherds of an earthenware stove for toasting sago at Kampung Pattimang Tua point to the importance of sago. Significant evidence for local ironworking was also found in the northwest sector of Kampung Pattimang Tua. Dammar gum has been excavated from Utti Batue. Finally, although our own excavations in Malangke have not recovered any gold, local farmers have reported that enormous quantities of gold were looted in Tampung Jawa.

Iron prills of unclear formation (Len Hogan, pers. comm.) have also been excavated at Baebunta's 15th-16th century palace center of Pinanto. However, it may be difficult to determine whether or not iron was imported from the upper Rongkong, because the road up to Limbong was washed away in the heavy 1998 rains, rendering excavations there logistically impractical. Iron samples which we collected on an earlier survey (Bulbeck and Prasetyo 1998) will be used for metallurgical analysis. Finally, Geoff Hope's recent inspection of Pattimang Tua indicated to him that it would have been an elevated, near-coastal location at the time of its occupation, subject only to occasional flooding.

Results: Further comment on ancient Luwu's economy would be inappropriate until the laboratory analysis of the excavated materials has been completed.

(4) The expansion of the Bugis as far as the head of the Gulf of Bone, which includes Ussu and Cerekang, did not occur until a comparatively late stage in Luwu's pre-Islamic history, if not indeed after Luwu's adoption of Islam. The coastal communities mentioned in Luwu's (presumably pre-Islamic) vassal list-Tampinna, Cerekang and Malili-were originally populated by non-Bugis speakers when they first fell beneath Luwu's authority.

Expectations: Pre-Islamic sites distributed around the head of the Gulf of Bone will produce little, if any, evidence of an ancient Bugis presence. There will be no convincing evidence of cremated human remains, this being the usual Bugis mortuary practice immediately prior to Islamization (e.g. Hadimuljono and Macknight 1983; Bulbeck 1996-7), or of earthenware pottery with identifiably Bugis designs or form.

Evidence: In immediately pre-Islamic times, cremation was clearly practiced by the Baebunta people, and also by the Wotu, according to reports of the contents found in buried martavans and associated grave goods by local looters. At Tambu-Tambu, near Wotu, this practice apparently dates as far back as the 13th-14th centuries. earlier this month recovered evidence. Hence, non-Bugis speakers in Luwu may have begun cremating their deceased even before the Bugis did. Nonetheless, pre-Islamic mortuary traditions other than cremation may still be taken as evidence against a Bugis affinity. This would appear to be the case east of Wotu, from Tampinna to Malili, where east-west inhumations, and storing uncremated human bones in or near jars, are the only pre-Islamic mortuary traditions for which we have positive evidence. However, we should point out that that investigations into the historical archaeology of Cerekang has been limited in the extreme, owing to the sacred status of Cerekang's historical sites. Study of the earthenware pottery (looking for identifiably non-Bugis designs or forms, which are absent from the Malangke sites) may provide more reliable information.

Results: From Wotu eastwards, with the possible exception of Cerekang, the spread of Bugis speakers appears to have occurred later than at Malangke.

(5) An industry collecting weapons-grade iron, smelting it and exporting it for the production of Java's renowned pamor luwu krisses had been established by the 14th century.

Expectations: The iron smelting industry at Matano, which is included in Luwu's circa A.D. 1500 vassal list (Caldwell 1995), will be found to date back to the 14th century, and to have produced hard, nickeliferous iron consistent with the quality of the pamor luwu renowned in Majapahit Java. Failing that, evidence for the 14th century production of weapons of Limbong iron ore (from Bukit Porreo' and Bukit Pangiwangen) will be found, either at Limbong sites themselves, or at Malangke coastal sites.

Evidence: Len Hogan has recently reported the trace presence of nickel in an iron prill from Rahampuu 1. He points out that smelting the ore should enrich the presence of nickel, so this prill could correspond to smelted iron of 1-10% nickel content. His analysis confirms the hypothesized production of nickeliferous iron at Matano. However, to judge from the several surveys of Matano by OXIS, and the four excavations, the earliest keramik is Sawankhalok in small quantities, probably no earlier than the 16th century Ming sherdage also recovered at low frequencies. The available evidence from Nuha, Sukoyu and Pontanoa Bangka suggests that settlement here, and any ironworking, occurred later than at Matano. We await the results from 11 charcoal samples, submitted for radiocarbon dating, from Rahampuu 1 and Pandai Besi, to determine whether or not iron smelting had commenced at Matano before the 16th century. Note that on a recent trip to the circa 12th-13th century site of Katue, we found a surface find provisionally identified by Geoff Hope as nickeliferous iron ore. If this can be confirmed in the laboratory, it would suggest at least some exportation of nickeliferous iron ore from somewhere near Lake Matano by the 14th century. (Traces of iron had been found very rarely during the Katue excavations.)

The possibility of a 14th century iron industry utilizing Limbong iron ore also remains possible. Detailed analysis of the ironworking levels at Kampung Pattimang Tua will be required before their precise dating, between the 14th and 16th centuries, can be deduced. Metallurgical analysis may discriminate the source of its iron, whether from Lake Matano or from Limbong.

Results: It is difficult to believe that the 14th century rise of Luwu, which appears to have caught the attention of the Javanese, was not somehow associated with Luwu's access to, and potential control over, the high-quality iron ore in Luwu's hinterland. However, the positive evidence for this proposition so far is irregular and no more than suggestive.